1 Nobo Jibon Final Report March 20, 2015 Save the Children Bangladesh Quantitative Evaluation Results:   Nobo Jibon Multi‐Year Assistance Program TANGO International Review Team:  Mark Langworthy – Team Leader  Thomas R. Bower  Golam Kabir  A K M Towfique Aziz 2 Table of Contents Table of Contents ............................................................................................................................ 2 List of Tables .................................................................................................................................. 4 List of Figures ................................................................................................................................. 7 Acknowledgements ......................................................................................................................... 8 Acronyms ........................................................................................................................................ 9 Executive Summary ...................................................................................................................... 11 Save the Children Bangladesh Quantitative Evaluation Results: Nobo Jibon Multi-Year Assistance Program ....................................................................................................................... 17 1. Introduction ........................................................................................................................... 17 2. Evaluation Methodology ....................................................................................................... 19 2.1 Methods for Endline QPE ................................................................................................... 19 A. Study Design ..................................................................................................................... 19 B. Sample Design .................................................................................................................. 20 C. Questionnaire .................................................................................................................... 23 D. Field Procedures ............................................................................................................... 24 E. Data Analysis .................................................................................................................... 26 2.2 Methods for Qualitative Study ............................................................................................ 33 A. Study Design and Objectives ............................................................................................ 33 B. Study Sample .................................................................................................................... 33 C. Instruments ........................................................................................................................ 34 D. Data Collection ................................................................................................................. 34 F. Estimation of Household Food Security Categories ......................................................... 35 2.3 Study Limitations and Issues Encountered ......................................................................... 37 3. Endline Evaluation Findings ................................................................................................ 39 Household Food Security and Vulnerability Status .................................................................. 39 Household Income and Expenditures ........................................................................................ 45 SO1 – Maternal and Child Health and Nutrition (MCHN) ....................................................... 48 Anthropometric Indicators ..................................................................................................... 48 Childhood Illness, Child Feeding Practices and Antenatal Care ........................................... 57 Water, Sanitation and Hygiene (WASH) .............................................................................. 67 3 Qualitative Information (SO1) ............................................................................................... 68 SO2 – Market-based Production and Income Generation ......................................................... 69 Agricultural Production and Marketing Practices ................................................................. 72 Qualitative Information (SO2) ............................................................................................... 82 SO3 – Disaster Risk Reduction ................................................................................................. 83 Qualitative Information (SO3) ............................................................................................... 86 Project Participation .................................................................................................................. 86 Vulnerable Groups .................................................................................................................... 95 Women’s Decision Making and Empowerment .................................................................... 95 Child Rights and Protection ................................................................................................... 98 4. Conclusions ......................................................................................................................... 101 5. Recommendations ............................................................................................................... 103 Annex 1: Focus Group Discussions ............................................................................................ 104 Annex 2: Mean Values and Confidence Intervals for Indicator Performance Tracking Table (IPTT) Indicators ........................................................................................................................ 106 Annex 3: Procedures for Computing Household Economic and Food Security Status Indicators ..................................................................................................................................................... 112 Annex 4: Results of Factor Analysis on Food Security Variables.............................................. 117 Annex 5: Nobo Jibon Baseline Survey Household Questionnaire ............................................. 120 Annex 6: Additional Quantitative Data ...................................................................................... 155 SO1 Tables .............................................................................................................................. 155 SO2 Tables .............................................................................................................................. 171 SO3 Tables .............................................................................................................................. 180 Annex 7: Terms of Reference, Baseline ..................................................................................... 182 Annex 8: Terms of Reference, Endline QPE .............................................................................. 188 4 List of Tables Table 1: Sample size by district....................................................................................................................................22 Table 2: Indicator definitions and calculation methods..............................................................................................28 Table 3: Food security variables at endline, by food security .....................................................................................36 Table 4: Selected household characteristics, baseline and endline survey rounds.....................................................37 Table 5: Program goal indicators, by district...............................................................................................................40 Table 6: Program goal indicators, by food security category ......................................................................................43 Table 7: Program goal indicators, by sex of head of household..................................................................................44 Table 8: Household income and expenditures (in Tk), by district ...............................................................................46 Table 9: Overall stunting, by age and food security category .....................................................................................49 Table 10: Severe stunting, by age and food security category....................................................................................51 Table 11: Overall underweight, by age and food security category............................................................................53 Table 12: Severe underweight, by age and food security category.............................................................................54 Table 13: Overall wasting, by age and food security category ....................................................................................55 Table 14: Severe wasting, by age and food security category.....................................................................................56 Table 15:  Incidence of child diarrhea, by food security category...............................................................................57 Table 16: Source of treatment for child diarrhea, by food security category .............................................................58 Table 17: Children with fever during the last two weeks, by food security category .................................................59 Table 18: Source of treatment for fever, by food security category ...........................................................................59 Table 19: Incidence of child cough/cold, by food security category ...........................................................................60 Table 20: Source of treatment for child cough/cold, by food security category.........................................................61 Table 21: Breastfeeding practices, by food security category.....................................................................................62 Table 22: Child feeding and care giving practices, by food security category.............................................................63 Table 23: Nutrient consumption among PLW, by food security category...................................................................65 Table 24: Attendance at antenatal care sessions, by food security category .............................................................66 Table 25: Percentage of children 12‐23 months who received Vitamin A supplementation, deworming treatment within last 6 months, by food security category .........................................................................................................67 Table 26: Caregiver hygiene practices, by food security category ..............................................................................67 Table 27: Economic and food access indicators, by food security category................................................................71 Table 28: Economic and food access indicators, by sex of head of household ...........................................................72 Table 29: Use of improved agricultural techniques, by food security category ..........................................................73 Table 30: Type of improved agricultural technique used............................................................................................74 Table 31: Types of buyers for agricultural product .....................................................................................................75 Table 32: Use of marketing practices, by food security category................................................................................76 Table 33: Source of agricultural inputs........................................................................................................................77 Table 34: Summary statistics for agriculture, by food security category ....................................................................78 Table 35: Khash land/water body use, by food security category...............................................................................79 Table 36: Household production, by food security category.......................................................................................81 Table 37: Household preparedness and impact of recent disaster, percentage by district........................................83 Table 38: Households participation in SO1, SO2 & SO3; by food security category....................................................89 Table 39: Key program goal indicators, by program participation ..............................................................................89 Table 40: SO2 impact indicators, by program participation........................................................................................91 Table 41: Childhood feeding practices, by program participation in SO1 ...................................................................92 5 Table 42: Use of improved agricultural techniques, by program participation in SO2 ...............................................92 Table 43: Use of improved agricultural techniques, by technique and program participation (endline) ...................93 Table 44: Household preparedness and impact of recent disaster, by program participation...................................94 Table 45: Women's income earning activities and decision making, by food security category ................................96 Table 46: Women’s decision making and empowerment, by participation................................................................97 Table 47. Overall and severe stunting, by age and district........................................................................................155 Table 48: : Overall and severe stunting, by age and child sex ...................................................................................156 Table 49.  Overall and severe underweight, by age and district ...............................................................................157 Table 50: : Overall and severe underweight, by age and child sex............................................................................158 Table 51.  Overall and severe wasting, by age and district........................................................................................159 Table 52: : Overall and severe wasting, by age and child sex....................................................................................160 Table 53: Overall and severe child malnutrition indicators, by program participation.............................................161 Table 54: Breastfeeding practices, by district............................................................................................................162 Table 55: Incidence of and source of treatment for child diarrhea, by district.........................................................162 Table 56: Children with fever during the last two weeks, by district ........................................................................163 Table 57: Incidence of and source of treatment for child cough/cold, by district ....................................................163 Table 58: Child feeding and care giving practices, by district....................................................................................164 Table 59: Nutrient consumption among PLW, by district..........................................................................................164 Table 60: Attendance of antenatal care sessions, by district ....................................................................................166 Table 61: Caregiver hygiene practices, by district.....................................................................................................166 Table 62: Percentage of children 12‐23 months who received Vitamin‐A supplementation, deworming treatment   w/in last 6 months, by district...................................................................................................................................167 Table 63: Main source of drinking water, by district.................................................................................................167 Table 64: Main source of drinking water, by food security category ........................................................................168 Table 65: Safety of tube well, by district ...................................................................................................................168 Table 66: Safety of tube well, by food security category ..........................................................................................168 Table 67: Water storage practices, by district...........................................................................................................169 Table 68: Water storage practices, by food security.................................................................................................169 Table 69: Type of latrine, by district..........................................................................................................................169 Table 70: Type of latrine, by food security ................................................................................................................170 Table 71: Economic and food access indicators, by district ......................................................................................171 Table 72: Household income and expenditures (in Tk), by food security category ..................................................171 Table 73: Household income and expenditures (in Tk), by sex of head of household..............................................172 Table 74: Household income and expenditures (in US$), by district, food security, sex of head of household .......173 Table 75: Use of improved agricultural techniques, by food security category ........................................................174 Table 76: Use of improved agricultural techniques, by district.................................................................................175 Table 77: Households receiving agricultural training, by district ..............................................................................175 Table 78: Types of buyers for agricultural product, by food security........................................................................175 Table 79: Types of buyers for agricultural product, by district..................................................................................176 Table 80: Use of marketing practices, by district ......................................................................................................176 Table 81: Use and source of agricultural inputs, by food security category..............................................................176 Table 82: Use and source of agricultural inputs, by district ......................................................................................177 Table 83: Household production , by district ............................................................................................................177 Table 84: Access to agricultural land and water, by district......................................................................................179 Table 85: Economic and food access indicators, by food security category, by sex of head of household (US $)....179 Table 86: Household preparedness and impact of recent disaster, by food security category ................................180 6 Table 87: Households who have received disaster preparedness training, by food security category.....................181 Table 88: Early warning for disasters, by food security.............................................................................................181 7 List of Figures Figure 1: Nobo Jibon Operational Area .......................................................................................................................18 Figure 2: % of overall stunted (HAZ<‐2SD) children age 6‐59 months ........................................................................40 Figure 3: Household Food Insecurity Access Scale (HFIAS), mean value (0‐100) ........................................................42 Figure 4: Coping Strategy Index (CSI), mean value (0‐100) .........................................................................................42 Figure 5: Monthly Expenditure (in Tk) Per Capita (deflated).......................................................................................45 Figure 6: % overall underweight (WAZ<‐2SD) children age 0‐59 months ...................................................................52 Figure 7: % of overall wasting (WHZ<‐2SD) children age 6‐59 months.......................................................................55 Figure 8: Infants/toddlers 6‐23 months who receive a minimally acceptable diet.....................................................62 Figure 9: %age of PLW that consume food rich in iron ...............................................................................................64 Figure 10: Household Dietary Diversity Score (HDDS).................................................................................................70 Figure 11:  % of HH adopting 3 or more improved practices ......................................................................................73 Figure 12: Source of agricultural training reported by households at endline, by food security category .................75 Figure 13: Households who received warning within 12 hours of the last disaster....................................................85 Figure 14: Households who received disaster preparedness training, by district.......................................................85 Figure 15: Households that sought shelter within 12 hours of the last disaster.........................................................86 Figure 16: Reported rights of children acknowledged by parents...............................................................................98 Figure 17: Percentage of responses to the question: "Is it wrong to hit children whenever they do something bad?" .....................................................................................................................................................................................99 Figure 18: Percentage of responses to the question: "What children should be protected from?".........................100 8 Acknowledgements TANGO International wishes to thank all of the individuals who helped support the Nobo Jibon endline evaluation. While it is not possible to exhaustively identify every individual involved, the team is particularly grateful to a number of Save the Children staff members for their efforts and contributions to the endline evaluation: Md. Shaidullah, Naim Abu, and in particular, Toufique Ahmed for helping coordinate and lead the evaluation preparation, training, and field work activities. Special thanks also goes to Maqbul Bhuiyan and his excellent team at Data Management Aid for their work during enumerator training and field work preparation which resulted in a timely and high-quality data collection effort. Thanks goes to Masud Rana for his quality translations and note taking. It should also be noted that the data collection would not have been possible without support from the highly dedicated Save the Children field staff that provided logistical support during evaluation field work, and more importantly, supported Nobo Jibon activities through the life of program. Finally, we are most indebted to the individuals and families who gave freely of their time and company to be interviewed by our teams. Without their generosity and openness in welcoming us into their homes and sharing invaluable information about their lives, this important evaluation would have never happened. TANGO International 21 March 2015 9 Acronyms ADPC Asian Disaster Preparedness Center ANC Antenatal care BCC Behavior Change Communication CC Community clinic CPP Cyclone Preparedness Program DHS Demographic and Health Survey EP Extreme poor EPI Expanded Program on Immunization FFP (Office for) Food for Peace FFW Food for Work FGD Focus Group Discussion GMP Growth monitoring and promotion GOB Government of Bangladesh HDDS Household Dietary Diversity Score HPP Homestead production poor HH Households IPTT Indicator Performance Tracking Table IR Intermediary result IYCF Infant and Young Child Feeding KAP Knowledge, attitudes and practices LOA Life of Award MAHFP Months of Adequate Household Food Provisioning M&E Monitoring and evaluation MCHN Maternal and Child Health and Nutrition MTR Mid-Term Review MYAP Multi-Year Assistance Program NGO Non-governmental organization NJ Nobo Jibon PLW Pregnant and lactating women PP Productive poor PPS Probability proportional to size QPE Quantitative Performance Evaluation SC Save the Children SO Strategic Objective SOW Scope of Work TBA Traditional birth attendant U2 Under two years of age U5 Under five years of age UDMC Union Disaster Management Committee UH & FWC Union Health and Family Welfare Center USAID United States Agency for International Development VDC Village Development Committee VDMC Village Disaster Management Committee VHC Village Health Committee 10 VSLA Village Savings and Loan Association WASH Water, Sanitation, and Hygiene WHO World Health Organization 11 Executive Summary Since June 2010, Save the Children has been implementing the United States Agency for International Development (USAID)-supported Title II PL480 Multi-Year Assistance Program (MYAP) in Bangladesh, “Nobo Jibon.” The program is designed “to reduce food insecurity and vulnerability for 191,000 direct beneficiary households…in ten upazilas of Barisal Division over five years.” Initially, there were 9 upazilas; however, there was an administrative restructuring over the course of the program and 1 upazila was divided into 2 separate upazilas, resulting in a total coverage of 10 upazilas by program end. It has three strategic objectives (SOs) in the areas of maternal and child health and nutrition (SO1), market-based production and income generation (SO2), and disaster risk reduction (SO3), as well as a cross-cutting gender component. The GOB fund also provided critical support to Nobo Jibon and was invaluable to the program. This report documents the findings of the program’s final quantitative performance evaluation (QPE), conducted November 2014 – January 2015 by TANGO International, Inc. The purpose of the final QPE is to measure changes in project impact and outcome indicators over the life of the Nobo Jibon project, in order to assess the extent to which project objectives have been achieved, measure the overall impacts on populations in the project areas, assess the assumed causal pathways linking project activities to outcomes and impacts, and determine how interventions contributed to achieving project goals. Another key function of the final QPE is to provide current status for key indicators included in Nobo Jibon’s Indicator Performance Tracking Table (IPTT). Context The food security situation in Bangladesh was volatile at the point of program inception in 2010. Despite real wage growth in the previous five years leading to program initiation, a high rate of households, 31.5 percent, were still in poverty. High food commodity prices, rising since 2007, exacerbated an already poor food security situation. Food insecurity at a national level was extremely high as measured by the Household Food Security Access Scale – at the beginning of 2011; the reported value was 69, a value more than double (2013: 33) what was reported nearly two years later at the end of 2013.1 Child feeding practices, maternal health, and child nutrition were persistent problems on a national level at program commencement. An alarming number of children, 41 percent as measured by the 2011 DHS survey, were stunted, 16 percent wasted, and 36 percent underweight. Only 21 percent of children age 6-23 months were fed appropriately based on 1 State of Food Security and Nutrition in Bangladesh: 2013. Food Security and Nutrition Surveillance Project (FSNSP), 2014. Helen Keller International and James P. Grant School of Public Health. 12 infant and young child feeding practices, over half of children 6-59 months were reported as anemic, and 42 percent of ever-married women age 15-49 were anemic as well.2 Methodology The Nobo Jibon QPE utilized an ‘adequacy design’, or non-experimental design for simple pre￾post comparison of results. The evaluation survey was population-based with the sample drawn randomly from the sample frame of all households residing within the action areas of Nobo Jibon. The sample size was determined to provide statistically representative results for indicators at the level of household and children under five years of age. A two-stage sample selection process was used to select households to be interviewed. In the first stage, a total of 62 clusters (villages) were selected in each of the three program districts. In the second stage, 30 households were interviewed in each of the selected villages. The households were selected from a census listing of all households in the selected villages. During analysis the sample was weighted to account for the fact that within the three districts, the proportion of sampled households to district population was different In addition to the quantitative household survey, a small qualitative study was also conducted. The purpose of this qualitative study was to provide complementary information from project participants about their perceptions of how they benefited from project interventions as well as well as their assessments of the strengths and weaknesses of project implementation strategies. Focus groups, 24 in total, were conducted from a sample of purposefully selected villages that included community groups and committees supported by Nobo Jibon SO1 and SO2 activities, as well as, villages considered at high-risk to disaster. Findings Comparison of baseline with endline values demonstrates that the Nobo Jibon project met or surpassed targets for all SO1 and SO2 impact indicators measuring household nutrition and food security status. Details of project indicators at baseline and endline as well as target values are provided in Annex 2. In particular, the endline values for all anthropometric indicators, Household Food Insecurity Access Scale (HFIAS), Coping Strategy Index (CSI), Household Dietary Diversity Score (HDDS), and Months of Adequate Household Food Provisioning (MAHFP) met the target values for these indicators. The results for the SO3 impact indicators are favorable, as well; the percent of households with plans to protect lives and assets increased nearly 40 percent from baseline to endline. Several other SO3 impact indicators, increased significantly, as well, including: percent of households that received warning within 12 hours of 2 Bangladesh Demographic and Health Survey (DHS): 2011. National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ICF International, 2013. 13 the last disaster, percent of households that received disaster preparedness training, and the percent of households that sought shelter following the last disaster.. SO1 Maternal Child Health and Nutrition (MCHN) Goal indicators and impact indicators3 for SO1, particularly anthropometric indicators, improved dramatically from baseline to endline. The prevalence of overall stunting for children aged 6-59 months declined 19 percent - from 44 percent at baseline to 35 percent at endline. This significantly exceeded the program target of 40 percent. This result is comparable to national statistics – stunting fell nationally from 45 percent in 2010 to 35 percent in 2013.4 Declines in the prevalence of underweight children (aged 0-59 months) and overall wasting (aged 6-59 months) were even more favorable over the life of program, 31 percent and 32 percent respectively. The endline results for underweight children 0-59 months (27 percent) and wasting of children 6-59 months (11 percent) also surpassed program targets of 36 percent and 14 percent, respectively. Reductions in underweight children and wasting compared quite favorably to national trends – underweight children remained flat at 32 percent from 2010 to 2013. National rates of child wasting rose from 10 percent in 2010 to 12 percent in 2013.5 Food security for the Nobo Jibon sample population improved markedly and beat targets over the life of the program, as measured by the HFIAS and CSI indices. The HFIAS index declined 32 percent and the CSI index declined 38 percent for all households sampled from baseline to endline. The HFIAS index at endline for all households sampled was 19 percent compared to a program target of 26 percent, while the CSI index at endline was 8 percent compared to a target of 12 percent. These improvements were supported by high rates of adoption of recommended practices for child feeding and care, diet and treatments for pregnant and lactating women. Infants and toddlers (aged 6-23 months) receiving a minimally acceptable diet increased from a baseline value of 6 percent to 23 percent of households surveyed at endline, although did not meet the program target of 25 percent. The prevalence of children (aged 12-23 months) receiving deworming treatment increased dramatically as well, 74 percent over the life of the program, and in this case was in line with the program target (33 percent endline; 30 percent target). Infant and mother health was supported by strong improvements in nutritional behaviors of pregnant and lactating mothers (PLW). PLW reporting consuming foods rich in iron and vitamin A increased 184 percent and 166 percent, respectively. At endline, 91 percent of mothers 3 See the IPTT table in Annex 2 for indicator types. 4 FSNSP, 2014. 5 FSNSP, 2014. 14 reported consuming iron rich food (60 percent target) and 60 percent reported consuming foods rich in Vitamin A (60 percent target). It is important to note that these changes in practices were observed for both respondents that participated in SO1 interventions and those that did not report participating directly in these interventions. These results suggest that Nobo Jibon has helped to contribute to a change in child care and nutrition practices, and household hygiene practices that has been also supported by the government and other organizations that have reached households not participating directly with Nobo Jibon, or that Nobo Jibon interventions have indirectly reached individuals in project areas that have not been participants in project activities. SO2 Market-based Production and Income Generation Goal and impact level indicators for SO2 have also improved substantially from baseline to endline. The HDDS increased 21 percent, to 5.7 at endline for all households surveyed exceeding the program target (target of 5.5). MAHFP increased from 9 months to 10 months over the life of the program, but fell short of the program target of 11 months. Livelihoods improved as measured by agricultural product sales. The average value of agricultural product sales (in Taka, real values adjusted for inflation), including livestock and crops, increased 11 percent to 11,646. Outcome indicators of adoption of recommended practices show large percentage increases from baseline to endline, but the overall levels are quite low even at baseline. For example, the percent of households adopting at least three improved production practices increased by over 40 percent from baseline, but the endline value is still less than seven percent of all surveyed households. The percent of households that have adopted improved marketing practices shows the same pattern of large percentage increase from a very low initial value, but a low actual value, less than two percent of all households, at endline. Optimistically, the large percentage increases in adoption of improved farming techniques and business practices are higher for SO2 participants compared to non-participants, implying that SO2 programming is affecting positive change in farmer behavior. These results suggest that there is interest on the part of farmers to adopt these practices, but there is probably continued need for promoting the messages to large numbers of farmers into the future. SO3 Disaster Risk Reduction Information about changes in disaster preparedness shows positive results. The percent of households reporting that they have plans to protect lives and assets in the event of a disaster increased from baseline to endline, 19 percent for households surveyed. Despite the impressive growth, the proportion of households with disaster preparedness plans (64 percent) fell short of the program target of 75 percent. The percentage of households reporting that they are able to resume livelihoods within two weeks after a disaster increased somewhat, eight percent over the life of the program, to 80 percent of all households at endline, also falling short of the program 15 target of 90 percent. For both these indicators, higher percentages of households that participated in SO3 activities reported positive responses than those that did not participate in SO3. Vulnerable groups One important thrust of the programming strategy of Nobo Jibon has been to reduce the exclusion of women and other vulnerable groups (especially children) from economic and social opportunities and to enhance the economic empowerment of women. According to information collected from women who had access to income, their economic empowerment, as measured by decision-making authority over income and economic activities, has increased from baseline to endline, although this change cannot necessarily be attributed with participation in project activities.6 The qualitative research suggests that the project interventions with youth seem to have a positive influence on the empowerment of girls and women. Potential implications from this research are that i) programming strategies directed toward youth may enhance the empowerment of women, and ii) indicators of empowerment should be measured on youth. Conclusions and Recommendations While many of the SO1 impact indicators, along with the childhood stunting goal indicator, improved dramatically over the life of the program, further analysis of achievement disaggregated by project participation showed that there was no significant differences in these measures between project participants and non-participants. A possible cause of these observed results is rooted in the range of government programs projects supported by non-governmental organizations that have been providing similar support and services to the rural poor in Bangladesh over many years. This is not to say that Nobo Jibon SO1 programming was not useful or effective, as it certainly was invaluable to the villages, households, mothers, and children that received program support. However, attribution of positive program effects is difficult when there are multiple programs, services, and messaging being delivered in the same geographic areas. One area where there was improvement that might be attributed to program participation was in farmer adoption of appropriate agricultural practices. While the percentage improvement in farmers adopting these improved practices was large (9.5 percent of SO2 participating farmers adopting vs. 5.9 percent of non-SO2 participating farmers adopting at endline), there are still an overwhelming proportion of farming households that could benefit from SO2-type programming support – even after strong growth, at endline only 6.9 percent of farmers had adopted 3 or more appropriate agricultural practices. 6 Following FFP guidance for performance monitoring evaluation design, the sample was not drawn such that statistically representative conclusions can be drawn between participant and non-participant households. See 2.3 Study Limitations. 16 Given the strong ongoing investment in health and nutrition programming by government and other private sector resources, taken together with strong gains in health and nutrition observed in the program area over the course of this evaluation, now might be an opportune time to perform a review of the mother, child, health, and nutrition (MCHN) programming being offered to find areas where there is overlap with complementary offerings by other organizations and/or the government of Bangladesh (GOB) and consolidate these services to eliminate any possible redundancies. Following this review, any resources liberated could be diverted towards the programming directed toward enhancing livelihoods that Nobo Jibon has demonstrated to be successful at effectuating positive change in farming practices. In the future, project M&E plans should include an integrated final project evaluation design that includes both qualitative and quantitative components. Ideally, monitoring and evaluation design of the next round of programming (or a separate impact evaluation) would incorporate testable hypotheses and a representative comparison group to evaluate the effectiveness of project activities for beneficiaries vs. non-beneficiaries. 17 Save the Children Bangladesh Quantitative Evaluation Results: Nobo Jibon Multi-Year Assistance Program 1. Introduction Program background Save the Children began implementing “Nobo Jibon” in Bangladesh in May 2010. The program is a USAID-supported $55.73 Million Title II PL480 Multi-Year Assistance Program (MYAP). TANGO International, Inc., a consulting firm based in Tucson, Arizona, USA, has been contracted to conduct the endline Quantitative Performance Evaluation (QPE) of the program. The main objective of the QPE is to review a) the achievements of the project relative to its prescribed targets, b) gauge whether the assumed relationship between project activities and outcomes and impact on communities is valid, and c) assess progress toward the overall goal of positive impact on food security of target communities. The overarching goal of the Nobo Jibon (NJ) program was to reduce food insecurity and vulnerability for 191,000 direct beneficiary households, or nearly one million people, in ten7 upazilas of Barisal Division over five years. Three strategic objectives (SOs) of the program aligned with USAID’s priorities for Bangladesh and with the Government of Bangladesh’s national health and food security policies. The Strategic Objectives of Nobo Jibon program include: • SO1: Maternal Child Health and Nutrition (MCHN) - Improved health and nutritional status of children under the age of 5 years (U5) and Pregnant and Lactating Women (PLW). • SO2: Market-based Production and Income Generation - Poor and extremely poor households have increased production and income. • SO3: Disaster Risk Reduction (DRR) - Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters. To maximize the impact of household food security, Nobo Jibon was designed such that a large proportion of households8 would participate in all three SOs. 7 See QPE Scope of Work (SOW) 8 72,000 households were targeted to participate in all 3 SOs (See QPE SOW). 18 Figure 1: Nobo Jibon Operational Area Endline Evaluation Objectives The endline study aims, through quantitative and qualitative surveys of a representative sample of households in the program impact area to review the project achievements relative to its targets and progress towards the overall goal. The purpose of the endline QPE is to assess the performance of key indicators against the baseline values to measure strategic objectives and intermediate results of Nobo Jibon. Specific objectives include:  To assess whether progress against agreed indicators/targets have met end of project benchmarks as documented in the indicator tracking table  To evaluate the theory of change through establishing plausible links between inputs, outputs, outcomes and impacts on target population  To determine whether critical strategies are missing that were needed to achieve Nobo Jibon’s goal; 19  To assess the overall impact of the project on target population; To identify where interventions, in isolation or in combination, were insufficient to meet program goals and, in cases where goals were not met, assess whether that was due to faulty logical reasoning/hypothesized causal pathways, to implementation shortcomings, or to other factors Endline information will be used to suggest design adjustments to improve the quality of future programming. Findings will also be used to identify where interventions were insufficient to meet program goals and, where goals were not met, assess whether that was due to faulty logical reasoning/ hypothesized causal pathways, to implementation shortcomings, or to other factors. 2. Evaluation Methodology 2.1 Methods for Endline QPE A. Study Design The QPE is principally a quantitative survey, supplemented by a limited qualitative study used to triangulate the results stemming from the quantitative household data. The population-based survey serving as the main component of the QPE is modeled after the corresponding baseline survey, thus allowing for the comparability of statistically representative results across survey rounds9 . The population-based survey used for the QPE, per FFP guidance, was structured after the corresponding baseline survey and includes structured questions related to relevant themes for all three strategic objectives. The data collected is used to estimate point prevalence and measure progress for key agriculture, nutrition, and gender-related indicators, including those contained in the program Indicator Progress Tracking Table (IPTT). Under the design of a population based survey, data were collected from both beneficiary and non-beneficiary households. Additional analysis has been included: comparisons of beneficiary households vs. non-beneficiary households, as well as, causal analysis exploring the relationships among project outcomes and higher level impacts. While the overall QPE is not designed to provide a clear counterfactual that can provide a direct measurement of project effects, the supplementary analysis is provided to provide preliminary indications about project effects that can be pursued more fully follow-on qualitative study and recommendations for further research that can inform the next round of program design. Overall, the surveys are consistent with the Office for Food for Peace (FFP) guidance for the design of program monitoring evaluations. As noted above, the surveys and analysis were kept 9 Nobo Jibon QPE Statement of Work (SOW) 20 as consistent as possible to allow for comparable results between the baseline and endline surveys. B. Sample Design The sample size was estimated based on the outcome indicator stunting among children 6-59 months. The indicator value and the design effect are obtained from the NJ baseline dataset. The FANTA Sampling Guidelines10 were used to calculate a sample size capable of detecting a 10 percent reduction in the child stunting indicator over the five-year intervention. The minimum sample size required per survey round was computed as follows: n = [(Zα + Zβ) 2 * {P1(1-P1) + P2(1-P2)}/(P2-P1) 2 ] * D * Nf where: n = required minimum sample size per survey round or comparison group (strata) P1 = stunting rate at baseline, 43.9% = 0.439 P2 = the expected level of stunting at endline for the program area such that the quantity (P2 - P1) is the size of the magnitude of change it is desired to be able to detect, NJ life of award (LOA) target, 39.5% = 0.395 Zα = the Z-score corresponding to the degree of confidence with which it is desired to be able to conclude that an observed change of magnitude (P2-P1) would not have occurred by chance (α - the level of statistical significance for one-tailed test), 95% = 1.645 Zβ = the z-score corresponding to the degree of confidence with which it is desired to be certain of detecting a change of magnitude (P2-P1) if one actually occurred (β - statistical power), 80% = 0.840. D = Actual NJ baseline design effect for stunting = 1.308 Nf = Non-response factor (assuming a 10%11 non-response rate) = 1.10 Based on these parameter values, the estimated sample size (n) was 2,034. Thus, the minimum required sample size per survey round for the entire program area is 2,034 children under five years of age (U5). Considering that not all households have U5 children, the sample size was adjusted according to the Addendum to Fanta Sampling Guide to ensure that a sufficient number of U5 children were measured12. Assuming that the proportion of households with U5 children is 50 percent and that the average number of U5s in the population is 11.5 percent and the average 10 Sampling Guideline, FANTA III, Robert Magnani, 1999 11 NJ baseline findings show less than 5% non-response rate, however given the change from using the random walk method, which produces lower non-response, to census lists the estimate was adjusted to the higher 10% non￾response rate. 12 Stukel, Diana & Deitchler, Megan. Addendum to FANTA Sampling Guide by Robert Magnani (1999): Correction to Section 3.3.1 Determining the Number of Households that need to be Contacted. March 2012. 21 households size is 4.413, the total number of households required to be interviewed to reach 2,034 U5s is 4,886 households.14 This sample size is adequate to detect a 10% reduction in the stunting rate of children U5 at the program level (LOA target in IPTT). In order to have comparable results with the baseline, the sample design of the baseline round was followed for the endline. The baseline sample size calculation was computed based on the following criteria: 1. The sample was powered to detect a 15 percent difference in stunting across comparison groups 2. The target number of households calculation was computed using an inflation factor based on the proportion of households with under 5s (45 percent) and the average number of under 5s per household (1.5) 3. The sample was stratified by district (3). 4. The design effect used for the baseline was from the endline survey of the previous MYAP (Jibon o Jibika) The minimum required sample size computed for the baseline was 5,082 households, larger than the minimum sample size to detect a 10 percent change in the prevalence of stunting at the project level, as described above. The computation of the minimum sample size for the endline was adjusted, based on the actual stunting rate and the actual design effect of the stunting rate from the baseline survey round. In addition, because the sampling of households was from census listing files rather than a random walk, the non-response rate was increased to 10 percent. The sample size was computed from the same formula: n = [(Zα + Zβ) 2 * {P1(1-P1) + P2(1-P2)}/(P2-P1) 2 ] * D * Nf where: n = required minimum sample size per survey round or comparison group (strata) P1 = stunting rate at baseline, 43.9% = 0.439 P2 = the expected level of stunting at endline for the program area such that the quantity (P2 - P1) is the size of the magnitude of change it is desired to be able to detect, NJ life of award (LOA) target, 37.3% = 0.373 Zα = the Z-score corresponding to the degree of confidence with which it is desired to be able to conclude that an observed change of magnitude (P2-P1) would not have occurred by chance (α - the level of statistical significance for one-tailed test), 95% = 1.645 13 From DHS 2011. 14 All U5s in a selected household were measured for anthropometric indicators. The estimate for the proportion of children U5 per household is consistent with the baseline sample and data from the most recent Demographic and Health Survey (DHS). 22 Zβ = the z-score corresponding to the degree of confidence with which it is desired to be certain of detecting a change of magnitude (P2-P1) if one actually occurred (β - statistical power), 80% = 0.840. D = Actual NJ baseline design effect for stunting = 1.308 Nf = Non-response factor (assuming a 10% non-response rate) = 1.10 Based on these parameter values, the estimated sample size (n) per comparison group is 984 U5 children. Considering that not all households have U5 children, the sample size was adjusted to ensure that a sufficient number of U5 children were measured. Assuming that the proportion of households with U5 children is 50 percent and that the average number of U5s per household is 0.5, the total number of households required to be interviewed to reach 984 U5s is 1,968 per stratum (district), or a total sample of 5,904 households in the three strata. For survey logistics reasons, the number of households to be surveyed per district was increased to 1,984. Table 3 shows the details about the sample size and again, is sufficient to detect a 10 percent reduction in stunting for the entire program area. Table 1: Sample size by district Program districts (strata) Sample Size Number of clusters Number of sample households (HH)/cluster Barisal 1,984 62 32 Barguna 1,984 62 32 Patuakhali 1,984 62 32 Total 5,952 186 -- Selection of clusters15 A two-stage sample selection process was used to select households to be interviewed. In the first stage, 62 clusters were selected in each of the three program districts. In the second stage, 32 households were selected randomly from the sampling frame to be interviewed in each of the selected clusters, to give a total of 1,984 households interviewed in each district. The sampling frame was constructed by conducting a census in all sample clusters. The selection of clusters was selected using probability-proportional-to-size (PPS). This ensures that all households within the districts have an equal chance of being selected.16 The listing of clusters was arranged by union and upazila in the PPS selection process, to ensure wide geographic coverage of the district in the cluster selection process. Sampling frame A complete sampling frame for all households in the selected clusters is required and was constructed by conducting a census17. The census enumerators made hand sketches of the 15 Cluster is defined as the NJ program villages. 16 In larger clusters, the chance that any single household will be selected is smaller, but this is offset because larger clusters have a greater chance of being selected in the PPS procedure. 17 In order to comply with recent FFP guidelines, a listing using the census method will be applied, although the random-walk method was used as part of the NJ baseline survey. 23 clusters to obtain the patterns of household distribution in rural settlements. Clusters are quite compact geographically, with houses clustered along rural roads and pathways. These characteristics made it possible for survey teams to quickly identify the boundaries of clusters and locate roads, paths, and pockets of settlements within the clusters. Another characteristic of most clusters in the program area is that they have a linear geographic layout, often following the line of roads, rivers, or canals. Each household’s location in a given cluster was plotted on the hand-sketched map and assigned a household identification number.18 Smart phones were used to collect the information, which facilitates the quick generation of a full list of households in the census. Nobo Jibon field staff conducted the census survey and household mapping after receiving training from DMA. To ensure quality and neutrality, as a first-level check, M&E Technical Officers for the ten upazilas randomly checked the authenticity of the census list. As a second￾level check, SCI-M&E staff double-checked the list and took corrective measures if required. SCI applied appropriate protocols to ensure that the listings and maps were accurate. Lastly, as a third-level check, DMA deployed a team to randomly check the SCI-supplied list and propose necessary corrections when required. SCI coordinated with the DMA team for final quality control. Selection of households Households were selected randomly using the census of beneficiary and non-beneficiary households. Households were sampled without replacement, as the minimum required sample calculated already included an adjustment for estimated non-response. The randomly selected households in a cluster were circled on the hand-sketched maps. The data collection teams moved from house to house, each team aiming to complete 30 household surveys per day. Selection of respondents The household head and spouse/adult household members were the main respondents for this survey. Most of the SO1 questions are related to health and hygiene, infant and young child feeding (IYCF) and child care practices; thus, mothers or caregivers of children U5 covered the majority of the questions for SO1. However, pregnant women were also interviewed if available. If multiple mothers were present in the household, the mother of the youngest child was interviewed, consistent with the protocol in the baseline survey. The household head or male respondent was also involved in the interview process, to provide basic information at the household level. The household member directly involved in SO2 activities was interviewed to collect farming and marketing-related information. C. Questionnaire 18 GPS coordinates will also be collected for every household in the cluster. 24 The quantitative endline survey used the same NJ questionnaire as the baseline, though it was revised to comply with recent FFP/FANTA guidance and NJ program data requirements (Annex￾5). The English questionnaire was translated to Bangla and both versions were available on the mobile devices used for quantitative data collection. The modules included in the QPE survey are as follows: 1. Household Member (roster) 2. Household Background Information 3. Agriculture 4. Natural Disaster Preparedness 5. Food Security 6. Safe Water, Sanitation, and Hygiene Practices 7. Mothers/Caregivers of Children Under 5 Years 8. Individual Child Related Questions 9. Child Rights and Protection Questions 10. Child Anthropometry D. Field Procedures Timeline The ex-post review was conducted in the period October 2012 – January 2013, including preparation, field work, analysis, and reporting. Field research was carried out in Barisal Division in two phases: a household survey was conducted by Save the Children (SC) in October 2012 and qualitative fieldwork was conducted by the mid-term review (MTR) team from 14 November to 9 December 2012. Training, Piloting and Pre-testing A six-day training, including one day for field testing and adjustment of tools, was conducted in Patuakhali. The training was a combined session that included field supervisors, enumerators, as well as, anthropometric enumerators. The following topics and activities were covered:  Brief program overview and the objectives of the surveys  Survey methodology – team composition, sampling, household selection process  Detailed discussion of the questionnaire form (question-by-question)  Practice administering the questionnaire using tablets (via role play/mock interviews)  Role play to show the technique of asking some sensitive questions  1-day anthropometric training session, including a standardization exercise 25 The training also included a 1-day field-test exercise (including both the full household survey and anthropometric measurements), where enumerators went to a nearby program area mouza (village) not selected in the sample and practiced implementation of the survey in a field-setting. The purpose of the field-test was to test the soundness of the questionnaire and to identify potential problem areas, such as skip patterns, translation issues, sequence of questions, question coding, instructions to enumerators, and identifying difficult or sensitive questions. Upon completion of the field-test, a debriefing session was held with enumerators and supervisors to address any issues which arose. Two training manuals were developed to support enumerators and supervisors in the field.  Supervisor Manual: The enumerator manual covered; roles and responsibilities, general interview guidance, privacy, ethics, interview techniques, tablet guidance, sampling protocol, quality control, editing of surveys,  Enumerator Manual: The supervisor manual covered; roles and responsibilities, sampling protocol, quality control, spot checks, logistical support, survey editing, and technical (uploading of data) support and troubleshooting. Supervisors were instructed to review specific questions, and series of questions, prior to uploading data to the server. In addition, supervisors completed a purposeful spot check each day – verifying enumerators were collecting accurate data. In addition to the supervisor quality control mechanisms, data was uploaded to TANGO frequently, often daily, throughout the course of data collection. TANGO reviewed the data and provided the field coordinators with feedback on data quality, survey progress, and highlighted specific issues to be discussed with identified enumerators. Fieldwork Android tablets (Google Nexus Tablets) were used for quantitative data collection, using ODK (Open Data Kit) software. The use of mobile devices and an electronic questionnaire allow for the integration of data validation rules and consistency checks as part of data collection. It also reduces data entry burden and supports data accuracy, as data is entered at the interviewer level. Every record was stored and uploaded to a cloud server utilizing the built-in internet connectivity of the devices. This allowed the data analysis team to review data consistency every day and ensure the data were ready for analysis as soon as one day after field data collection was completed. The team leaders were responsible for re-interviewing up to two households per day using tablets. Team leaders also verified that non-response households were unavailable or truly opted out. The database software allows for the cross-referencing of re-interview records with the original records collected by the enumerators. At the end of each day, district coordinators reviewed the full electronic dataset collected. He/she ran data frequencies and cross-tabulations to verify data consistency at the interviewer level by comparing the re-interview data with the corresponding interview data. The district coordinator discussed discrepancies with the 26 concerned enumerator and the respective team leader to determine the reason for the discrepancies. The team leader followed up with appropriate measures to correct any deficiencies discovered. SCI representatives also traveled to the field to observe data collection by occasionally sitting in on interviews, reviewing the questionnaires, and speaking with enumerators and supervisors. One TANGO staff member involved in the entire process spent time in the field during the first week of data collection to monitor whether the data collection teams were collecting information appropriately. He provided immediate feedback and technical support as needed. This TANGO staff member also continued to monitor data consistency throughout the ongoing data collection process. Data Entry and Processing The ODK dataset (CSV format) was converted into an SPSS (Version 20) database for data management and analysis. Validated data were transferred to the main SPSS database daily. TANGO applied a comprehensive data analysis and tabulation plan according to the IPTT and baseline report prior to the data analysis stage. The ODK CAPI software included automated validation and consistency checks as part of the electronic survey. Examples include: responses for assets, income, and expenditures were constrained such that values were kept within reasonable ranges; children’s weight and height measurements were constrained to remain within minimums and maximums established as part of WHO guidelines; where applicable for multiple response questions, “Don’t Know” was not allowed as a response when other affirmative responses were selected; among many other automated constraints. The automated consistency checks programmed directly into the survey limited data entry errors associated with invalid and consistent data, as well as, outliers. SPSS statistical software was used to analyze the dataset, and World Health Organization (WHO) Anthro software was used for anthropometric data analysis. Syntax files were created to compute indicator and sub-indicator values. The analysis includes mostly descriptive statistics with some statistical hypothesis testing. Due to stratification, normalized sampling weights were used to adjust indicator value estimates. Also, complex analysis was performed to estimate standard error and confidence interval through the adjustment of the design effect. Missing data points were excluded from the denominator and the numerator for calculation of all indicators and descriptive statistics. Responses of “Don’t Know” were recoded to “null” values and included in the denominator. As an example, a question may contain response codes of “Yes”, “No”, and “Don’t Know”. All three responses are counted in the denominator, but only “Yes” may be counted in the numerator (unless the number of “Don’t Know” cases was sufficiently high to report). E. Data Analysis Sampling Weights 27 The Nobo Jibon endline survey sample was drawn with two-stage, stratified cluster sampling based off a sample frame generated by a separate household listing exercise. Clusters were equally allocated among districts. At the first stage, a sample cluster was selected independently with probability proportional to the cluster’s population in each stratum. The strata were the three districts encompassing the program area – Barisal, Barguna, and Patuakhali. The unequal probabilities of selection across strata caused by the equal number of clusters in each stratum were adjusted relative to the population of each stratum. Design weights were calculated based on the separate sampling probabilities for each sampling stage and for each cluster. The design weights are the reciprocal of each unit’s probability of selection into the sample: w௜,ௗ௘௦௜௚௡ ൌ 1 ௜݌ , where pi is the probability of selection, and where i denotes strata i=1,2,3. The sampling weight was calculated with the design weight corrected for non-response for each of the selected clusters. The household respondent weights are constructed by first calculating the design weight (w) and response rate (rr) for households in each stratum as follows. ݅ ݉ݑݐܽݎݐݏ ݅݊ ݏ݈݀݋݄݁ݏݑ݋݄ ݂݋ ݊݋݅ݐݎ݋݌݋ݎܲ ൌ௜௚௡ ௦௘ௗ,୦୦,௜w ݅ ݉ݑݐܽݎݐݏ ݅݊ ݏ݈݀݋݄݁ݏݑ݋݄ ݂݋ ݈݁݌ܽ݉ݏ ݈ܽ݊݊݁݀݌ ݂݋ ݊݋݅ݐݎ݋݌݋ݎܲ Response rates were calculated at cluster level as ratios of the number of interviewed households over the number of eligible households. ݅ ݉ݑݐܽݎݐݏ ݅݊ ݏ݁ݎ݊݊ܽ݅݋݅ݐݏ݁ݑݍ ݁݀ݐ݈݁݌݉݋ܿ ݏ݈݀݋݄݁ݏݑ݋݄ ݂݋ # ൌ ݅,݄݄ݎݎ ݅ ݉ݑݐܽݎݐݏ ݅݊ ݈݁݀݌ܽ݉ݏ ݏݐ݊݀݁݊݋݌ݏ݁ݎ݂݋ # The non-response adjusted sampling weight is the design weight divided by the response rate: ݅݃݊ݏ݅,݄݄,݀݁ݓ ൌ ݎݎ݅,݄݄,w ൌ ݐ݄݁݅݃ݓ ݈݅݊݃݌ܽ݉ݏ ݈݀݋݄݁ݏݑ݋ܪ ݅,݄݄ݎݎ . A separate sampling weight was further adjusted and applied to reflect households that have more than one mother or caregiver of children under 5. In households that included more than one mother/caregiver, only one mother/caregiver was interviewed, therefore a correction was applied to the sampling weight to reflect the differing probability of any given mother/caregiver being interviewed. 1 ൌ ܾ݅,݄݄,݉݋ݎ݌ ݅ ݉ݑݐܽݎݐݏ ݅݊ ݈݁݀݌ܽ݉ݏ ݈݀݋݄݁ݏݑ݋݄ ݐ݊݀݁݊݋݌ݏ݁ݎ ݅݊ ݏݎ݁ݒ݁݃݅ݎܿܽ ݎ݋ ݏݎ݄݁ݐ݋݉ ݂݋ # 28 The sampling weight applied to mother/caregiver respondents was calculated by dividing the household design weight by the mother/caregiver response rate. ݎݎ݅,݄݄,ݓ ൌ ,݉ݎݎ݅,݄݄,w ൌ ݐ݄݁݅݃ݓ ݈݅݊݃݌ܽ݉ݏ ݎ݄݁ݐ݋ܯ ܾ݅,݄݄,݉݋ݎ݌ Indicator Definitions and Tabulations Table 2 presents program indicators for which baseline information was collected. Table 2: Indicator definitions and calculation methods Indicator Type of respondents Main Disaggregation Method Impact indicators % children between 6 and 59 months stunted (height-for-age) Children 6-59 months Boy, Girl, <-2SD, <-3SD Calculate height-for-age z-score (<-2SD and <- 3SD) using new WHO/CDC standard over total number of children 6-59 months Average HH Food Insecurity Access Scale score HH Head/ Female HH member No disaggregation Calculate using FANTA guideline for “Household Food Insecurity Access Scale (HFIAS) for Measurement of Food Access: Indicator Guide” Average HH Coping Strategy Index (CSI) HH Head/ Female HH member No disaggregation The coping CSI measures household vulnerability to food insecurity in times of stress. The CSI is calibrated so that the maximum possible value is 100. A zero value indicates high food security (no coping strategies were used), and a value of 100 indicates extreme food insecurity. Calculate using guidelines in “ Measuring food insecurity: Can an indicator based on localized coping behaviors be used to compare across contexts?” by Maxwell, Daniel, Richard Caldwell and Mark Langworthy, Food Policy, Volume 33, Issue 6, December 2008. SO1 MCHN: Improved health and nutritional status of children U5 and pregnant and lactating women (PLW) Percentage of underweight (WAZ<-2) children aged 0-59 months Children 0-59 months Boy, Girl, <-2SD, <-3SD Calculate weight-for-age z-score (<-2SD and <- 3SD) using new WHO/CDC standard over total number of children 0-59 months Percentage of wasted (WHZ<-2) children aged 6-59 months Children 6-59 months Boy, Girl, <-2SD, <-3SD Calculate weight-for-Height z-score (<-2SD and <- 3SD) using new WHO/CDC standard over total number of children 6-59 months 29 Table 2: Indicator definitions and calculation methods Indicator Type of respondents Main Disaggregation Method % children 0-6 months exclusively breastfed Mother/ caregiver of children <2 years No disaggregation Exclusive breastfeeding refers to children up to six months of age who are given nothing but breast milk in the 24 hours preceding the interview divided by the number of children 0-6 months % of children 6-23 months of age who receive a minimum acceptable diet Mother/ caregiver of children <2 years No disaggregation This is a composite indicator of IYCF practices. The indicator gives an overall measure of the degree to which women have complied with the recommendation that infants age 6-23 months receive appropriate and adequate complementary foods in addition to breastmilk. IYCF feeding practices will be disaggregated by age group to estimate age-specific feeding practices. Calculation: no. of children 6-23 months who received solid, semi-solid or soft foods in addition to breastmilk during the previous day divided by total no. of children 6-23 months. Calculate per WHO 2008 IYCF guideline. % of caregivers demonstrating proper personal hygiene behaviors Mother/ caregiver of children <5 years No disaggregation “Proper personal hygiene behavior” refers to includes two dimensions: critical times and technique: Critical times for handwashing: After defecation. After cleaning babies’ bottoms. Before food preparation. Before eating. Before feeding children. Handwashing technique: Uses water. Uses soap or ash. Washes both hands. Rubs hands together at least three times. Dries hands hygienically – by air￾drying or using a clean cloth. According to FANTA guidelines, mothers/caregivers practice eight or more of the 10 practices listed are considered as practicing appropriate handwashing. % of beneficiary caregivers demonstrating food hygiene behaviors Mother/ caregiver of children <5 years No disaggregation “Food hygiene behavior” is achieved if the beneficiary caregivers practice all of the following: 1) Wash hands before food preparation 2) wash hands before feeding child 3) keep food covered. % of PLW who consume food rich in iron PLW No disaggregation Defined as pregnant and lactating women’s consumption of local iron-rich food within the last 24 hours. The locally identified iron-rich food/food groups are dark green leafy vegetables, fish, poultry, meat/offal/organs, and pulse/peanuts/ beans/ ground-nuts. 30 Table 2: Indicator definitions and calculation methods Indicator Type of respondents Main Disaggregation Method % of PLW who consume food rich in Vitamin A PLW No disaggregation Defined as pregnant and lactating women’s consumption of local Vitamin-A- rich food within last 24 hours. The locally identified Vitamin-A￾rich food/food groups are milk/dairy products, oil/fats/butter, mango/papaya/orange/jack-fruit, DGLV, carrots/pumpkins, egg. % of PLW who consume food rich in Calcium PLW No disaggregation Defined as pregnant and lactating women’s consumption of local calcium- rich food within last 24 hours. The locally identified calcium rich food/food groups are milk/dairy products. % of PLW taking iron or iron folate supplements in the last 7 day PLW No disaggregation Defined as pregnant or lactating women who took an iron folate tablet/ supplement within the last seven days. % of children 12-23 months who received Vitamin-A supplementation in the past 6 months Mother/ caregiver of children 12-23 months No disaggregation Children 6-59 months of age are supposed to receive a Vitamin-A capsule every six months from a regular Expanded Program of Immunization (EPI) session or Vitamin-A-plus campaign as supplementation. Accounting for the initial six months, he program will track Vitamin-A supplementation for children 12-23 months. % of mothers of children aged 6- 23 months who received high￾dose Vitamin A supplement within 8 weeks postpartum (6 weeks if not exclusively breastfeeding) in last pregnancy Mother of children 6- 23 months No disaggregation Every mother should receive one dose of Vitamin A within six weeks of delivery (postpartum). The mother of the child 6-23 months who received Vitamin-A supplementation within six weeks of delivery in her last pregnancy will be counted for this indicator. % of mothers attended Antenatal Care (ANC) session at least 4 times during last pregnancy PLW No disaggregation If a pregnant woman attends ANC sessions at least four times during pregnancy she will receive all program messages related to pregnancy and newborn/infant care. The monthly attendance of pregnant women at ANC sessions is important to ensure full ANC services. Calculation: No. of pregnant women who have attended ANC sessions at least four times, over total # PLW (over the defined period). % of beneficiary children 12-24 months receiving de-worming medication in previous 6 months Mother/caregiver of children 12-23 months No disaggregation Children 12-59 months of age are supposed to receive deworming tablet every 6 months from regular EPI session or Vita-A plus campaign as medication. Children 6-23 months are the direct beneficiaries. So the program will track deworming tablet receiving status of children 12-23 months through regular monitoring.19 % of beneficiary women whose husband attends ANC/PNC with her PLW No disaggregation This indicator will measure the extent of male involvement in maternal health care. SO2 Market-based Production and Income Generation: Poor and extremely poor households have increased production and income 19 The original indicator statement is “12-24” but it should be 12-23: in the baseline, data were collected for children 12-23 months, and the program continued to track for that age range. 31 Table 2: Indicator definitions and calculation methods Indicator Type of respondents Main Disaggregation Method Average HH dietary diversity score (HDDS) Female HH member (who cooks food) No disaggregation Dietary diversity score (DDS) does not measure dietary quality or calorie intake; it is a proxy for the socioeconomic status of the HH. HHs that consume more diversified food/food groups are considered to have a better economic status in terms of food security. Household dietary diversity is defined as the number of unique foods consumed by household members over a given period. The following 12 food groups are used to calculate the HDDS: 1. cereals 2. roots and tubers 3. pulses/legumes 4. milk and milk products 5. eggs 6. 6.meat and offal 7. fish and seafood 8. oil/fats 9. sugar/honey 10. fruits 11. vegetables 12. others (spices, sodas, etc.) This indicator is calculated using 24-hours recall: the respondent is asked “Yesterday, did you or anyone in your household consume (list of food groups). The sum of the “Yes” (Yes=1, No=0) responses is the score per household; an average score is calculated for the sample. Average number of months of adequate household food provisioning (MAHFP) HH Head/ Adult Female HH member No disaggregation The average number of months beneficiaries are able to meet their basic food needs. The indicator focuses on the desired outcome of improved food access. Food access depends on the ability of households to obtain food from their own production, stocks, purchases, gathering, or food transfers from relatives, members of the community, the government, or donors. A household’s access to food also depends on the resources available to individual household members and the steps they must take to obtain those resources, particularly exchange of other goods and services. The survey question for this indicator is, "Which were the months (in the past 12 months) in which you did not have enough food to meet your family’s needs?". % of HHs reporting increase in production of one or more products Farming HH member No disaggregation "Production" is defined as the food produced from the vegetable garden. "Increase" is defined as at least a 20% increase from the baseline. Average annual income from sale of agricultural products HH Head/ farming HH member No disaggregation "Income" is defined as net income from agricultural products. This information will be collected semi-annually; and averaged annually. 32 Table 2: Indicator definitions and calculation methods Indicator Type of respondents Main Disaggregation Method % of beneficiaries (farmers) using 3 or more sustainable/improved production practices. HH Head/ farming HH member No disaggregation The project will promote the following seven sustainable/ improved production practices: (use of) animal manure; compost; crop rotation; biological/organic pest control; mechanical pest control; integrated pest management; and treadle pump/drip irrigation/mobile pump. Those beneficiaries who practice at least 3 out of the 7 improved practices will be counted for this indicator. % of targeted Productive Poor (PP) HHs adopting improved marketing practices HH Head/ farming HH member No disaggregation “Improved marketing practices” are defined based on three criteria: (presence of) business plan (crop season, improved variety and market demand); bulking products (bulking and selling collectively through group); and high-value marketplace. The farmer HHs who practice these three things will be considered as "adopting" improved marketing practices. SO3 DRR: Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters % of HHs with a feasible plan to protect human life and productive assets during disaster HH Head/Adult HH member No disaggregation A HH is considered to have a “feasible plan” when HH members have a plan for evacuating vulnerable HH members, visit the shelter center in normal times, identify a safe shelter center, have a plan for dry food, and have a plan to protect livestock and other valuable assets. %of HHs able to resume livelihood activities within two weeks following a natural disaster. HH Head/Adult HH member No disaggregation This indicator will be reported if any disaster takes place after the baseline survey. "Resume livelihood activities" is defined as when HH members start their normal livelihood activities – earning income, farming, doing agricultural activities, doing household chores, etc. % of HHs that received location￾specific cyclone warning signal with adequate lead time HH Head/Adult HH member No disaggregation The definition of “adequate lead time” varies depending on signal level. The current government signal system is based on two ports: Mongla and Chittagong. Nobo Jibon is working with Asian Disaster Preparedness Center (ADPC) to develop a localized (union-level) early warning system. The project collects signal -specific early warning information will be collected during annual monitoring. Reporting of Results The analysis presented in this report includes two types of cross-tabulations for all project indicators: by district and by household food security category (terciles of low, medium, and high food security). All indicators are broken down by these categories, either in tables within the report narrative or in Annex 6. In addition to these breakdowns, some key indicators are also broken down by sex of household head, and by categories of household participation in project interventions. 33 Throughout this report, baseline values of selected program indicators shown in Table 1 are computed as the mean values of the overall sample. Mean values and 95% confidence intervals of all IPTT indicator variables at the total sample level are provided in Annex 2. Data presented throughout the report is coded to indicate significant differences. The significance, which statistical tests produced, is referred to as the p-value (probability value). The p-value can be interpreted as the probability of a difference occurring by chance alone. If all other biases are eliminated or accounted for, then one can assume that when this p-value is small, the differences are due to a factor other than chance. * p < 0.1 *Mean value is different between groups at the .10 significance level. All monetary indicators are converted from nominal values to inflation-adjusted values based on 2010 price index level, in order to permit direct comparability between baseline and endline values. The adjustment is based on the Bangladesh Consumer Price Index (CPI) reported by the World Bank. 2.2 Methods for Qualitative Study A. Study Design and Objectives In order to obtain other qualitative information about beneficiary perceptions of program activities, change in practices, stakeholder coordination and linkages to services, three qualitative evaluators spent four days in the districts to conduct beneficiary focus groups. The team used the qualitative information to inform the interpretation of program impact and outcome data obtained from the quantitative data collection process. B. Study Sample The qualitative study sample was drawn from the villages selected for the quantitative portion of the evaluation. All three districts encompassing the program area, Barguna (1 village), Barisal (1 village), and Patukhali (2 villages), were included in the sample. Villages were purposively selected to include those that are considered at high risk to disaster (e.g. that received SO3 training), as well as, those that included community groups and committees targeted by Nobo Jibon SO1 and SO2 activities, as outlined below. The qualitative team conducted 24 focus groups, as follows:  MCHN – PLW (8-12) (two groups)  MCHN – Adolescents (two groups) 34  MCHN – Village Health Committee (VHC; two groups)  MCHN – Fathers (two groups)  Livelihoods – Extreme Poor (two groups)  Livelihoods – Productive Poor (three groups)  Livelihoods – Women (two groups)  Disaster Management – Men (two groups)  Disaster Management – Village Disaster Management Committee (VDMC; two groups)  Disaster Management – Women (two groups)  Disaster Management – Youth Volunteers (two groups)  Disaster Management – Union DMC (one group) Some focus groups were separated by sex and some were mixed. In total, the groups included 153 women, 68 men, 18 girls and 15 boys. Annex 1 contains focus group details. C. Instruments The qualitative team used topical outlines to guide the focus group discussions. For each strategic objective, the teams explored the following general topics:  Participation (frequency of participation, m/f ratio, adolescents: how selected  Topics learned and relative importance  Changes in practices (noting gender differences)  Reasons for not changing practices (noting gender differences  Suggestions/recommendations (e.g., ways to enhance inclusiveness)  Sustainability Village committees were asked about the following topics:  Structure of committee  Responsibilities and activities  Interactions with community  Types of support received by NJ  Participation of women in the committees  Sustainability of the committees D. Data Collection The qualitative component of the evaluation was conducted by one international consultant and two local consultants with relevant specializations in food and livelihood security, health and nutrition, disaster risk reduction and adaptation, program management, commodity management, 35 and gender and governance. The international consultant is from the United States and worked in tandem with one of the two local consultants when translation was required. The team collected qualitative data from upazilas in each district as follows:  Barguna: Amtali upazila (one union)  Patuakhali: Dashmina and Galachipa upazilas (four unions)  Barisal: Barisal Sadar upazila (one union) In total, the qualitative team visited six villages in the areas listed above. They applied the instruments described in the previous section. F. Estimation of Household Food Security Categories The evaluation team used factor analysis to construct a proxy indicator of household food security based on a composite of a number of measured household characteristics of household economic status and food security indicators. Factor analysis enables identification of unique factors that summarize several dimensions of the food security status of households. Results (provided in Annex 4) from the factor analysis were used to identify and compare three distinct levels of food security status among sample households. The computed values of the principal component (component 1) were first ranked and then divided into terciles (three groups with an equal number of cases). These categories represent three levels of food security status among sample households. The elements included in the factor analysis were:  Household size  Per capita expenditures  Per capita asset index  Share of household expenditures spent on food  Household Dietary Diversity Score (HDDS)  Months of adequate household food provisions (MAHFP)  Household Food Insecurity Access Scale (HFIAS) score  Coping Strategies Index (CSI) These elements were pre-identified as household and livelihood characteristics that exogenously explain and are correlated with household food security. Annex 4 includes results from and a detailed explanation of how the factor analysis was used to construct the food security index. Table 3 presents data on these indicators of vulnerability, disaggregated according to food security status. By identifying the index scores of households in different food security 36 categories, the Nobo Jibon endline QPE provides a useful tool for measuring the impact of Nobo Jibon on highly food insecure and less food insecure households in the program area. Table 3: Food security variables at endline, by food security    Food Security Category Total    Lowest Middle Highest Sample Variables included in food security categorization Mean value Household size* 4.5 4.7 5.3 4.8 Per capita expenditures (TK/month)** 1,425 1,460 2,308 1,728 Per capita asset index* 43.5 63.3 100.7 69 Food share (%) of total expenditures * 59.5 57.3 45.8 54.2 Household Dietary Diversity Score (HDDS)* 4.7 5.4 7 5.7 Months of Adequate Household Food Provisions (MAHFP)* 8.4 10.9 11.9 10.4 Household Food Insecurity Access Scale (HFIAS)* 47.2 10.2 0.8 19.4 Coping Strategy Index** 20.9 3.7 0.4 69.1 Food Security Index (Mean) ‐1.0 ‐0.3 1.2 0.0 Food Security Index (Standard Deviation) 0.3 0.3 0.7 1.0 N 1,778 1,779 1,779 5,336 Note:  All food security categories are statistically significantly different from one another at the 10% (*).  High is statistically significantly different from low and middle at the 10% (**) Across the entire sample, household size, per capita expenditures, per capita asset index, dietary diversity, and months of adequate household food provisions increase as food security status increased. Per capita expenditures ranged from a high of Tk 3,323 among the most food secure households to a low of 2,025 among the least food secure households. Households in the low food security category also spend the most on food as a share of total expenditures (60 percent) compared to households in the medium (57 percent) and high (46 percent) categories Notable differences between categories were seen in HFIAS and CSI: the lowest food security households scored 47.2 on the HFIAS and 43.6 on the CSI, compared to 0.8 on the HFIAS and 100.4 on the CSI for the most food secure households. It should be noted that not every household is included in the food security terciles. In cases where households did not provide a response necessary to calculate one of the underlying variables included in the factor analysis described above, those households were excluded from the factor analysis (“missing values”). Therefore, the total sample size will differ when disaggregating by food security category, as compared to other disaggregations presented in this report such as district and/or sex of head of household. For this same reason, the baseline values 37 presented in this report for indicators disaggregated by food security category may differ slightly from those included in the IPTT table (Annex 2), due to missing cases in the food security index. 2.3 Study Limitations and Issues Encountered One potential limitation of the evaluation was the difference in evaluation design with respect to sampling between baseline and endline. At baseline, detailed household listings were unavailable; therefore, second-stage selection of households was conducted using the random walk method. At endline, a household listing exercise was conducted prior to the commencement of field work and households for the second-stage of sampling were chosen from among the household lists. When possible, sample selection from household listings is preferable. There are drawbacks to using a random walk for household sample selection, as opposed to household listings, the biggest being the potential for selection bias. Table 4 above includes general household characteristics that are expected to remain relatively constant over time, for both the baseline and endline samples. These characteristics include asset ownership, prevalence of farming as an income earning activity, and prevalence of other-income earning activities, such as wage labor, and rickshaw driving that might be indicative of lack of access to farming activities. Across the sample, several characteristics change significantly. More than two-thirds of households (68 percent) owned cultivable land, compared to 60 percent at baseline. Average farmland area increased 67 percent from 52 decimals at baseline to 87 decimals. In addition, the proportion of households with access to water bodies grew 26 percent to 81 percent of all households. Table 4: Selected household characteristics, baseline and endline survey rounds   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH that own cultivable land 59.2 67.8 14.5 * 5,024 5,345 Average farmland area (decimals) 52.0 86.9 67.1 * 5,026 5,346 % HH with access to water bodies 64.1 80.5 25.6 * 5,022 5,345 Average # cows 0.9 1.1 22.2 * 5,026 5,346 Average # goats 0.3 0.3 0.0 5,026 5,346 % HH primary occupation: day labor 20.7 18.8 ‐9.2 * 5,025 5,337 % HH primary occupation: rickshaw puller/boatman 5.9 5.2 ‐11.9    5,025 5,337 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 38 If the random walk sample selection technique produced a biased sample, one might expect to see several of the household characteristics to be different for the sample at endline compared to baseline. This was, in fact, true. The percentage of households with access to farmland and the average size of agricultural land owned are considerably higher at endline relative to baseline. This is also true with respect to ownership of large livestock (cows/buffaloes), while ownership of smaller livestock (goats/sheep) was relatively unchanged. The percentage of households whose primary income was derived from wage labor or rickshaw driving declined slightly from baseline to endline. This is contrary to what one would expect to see if the baseline sample was biased towards wealthier households. However, it could just suggest that the baseline sample was biased towards poorer households. Another explanation for the differences observed in these household characteristics is that there was a generalized, upward trend in these variables between the two survey rounds, either the result of project activities or external factors. Unfortunately, without additional information to determine if the observed changes are due to selection bias or underlying structural changes of household conditions. Another limitation of the QPE conducted for the Nobo Jibon project is that the study includes only a very small and limited qualitative component. This is because FFP made the decision to undertake a comprehensive qualitative evaluation of all three Title II projects in Bangladesh, and asked the awardees to conduct quantitative evaluations only, to measure changes in project indicators. In the Scope of Work, Save the Children requested a small qualitative component, to serve as a means of triangulation and verification of the quantitative results. Because of the limited scope of the qualitative component, only a small number of interviews could be conducted with focus groups of project participants and with key informants. The limited scoped of the qualitative component did not permit wider ranging interviews with other project stakeholders to get information about project implementation. As a result, the qualitative component of this QPE is very narrowly directed toward collecting information from a small number of project beneficiaries about their perceptions of project interventions. Regarding the minimum required sample size calculation and corresponding statistical power associated with the evaluation, the initial intention during the evaluation design phase was to allow for statistically valid comparisons for outcome indicators between the 3 districts (Barguna, Barisal, and Patuakhali). The comparisons per district were requested by Nobo Jibon and based on detecting a 15 percent difference in stunting in children under 5. At the time of baseline, this sample size was more than sufficient to capture a sufficient number of children under 5 to detect a 10 percent reduction in stunting for the whole program area, per FFP requirements and consistent with the program target. There was no attempt, at endline, to adjust the sample size in an attempt to power the evaluation such that statistically valid comparisons could be made for child stunting between districts. This simply was not feasible from a budgetary perspective. The decision was made to maintain consistency from the baseline with respect to the sample size calculation at endline, as this 39 minimum required sample was assumed to be more than sufficient to detect a 10 percent reduction in stunting for the whole program area. This in fact, was the case. Granted, it should be noted that any district comparisons made in this report will not be statistically valid for the anthropometric indicators. Finally, it should be noted that in following FFP guidance for performance monitoring evaluation design (as opposed to for an IE), a statistically representative comparison (or control) group was not built into the evaluation design. However, the population based survey design did include a large proportion of households from program villages that did not participate directly in Nobo Jibon activities, from which a limited amount of analysis is included in this report, comparing non-participant households to participant households for certain key indicators. While the analysis is constructive, it is only meant to provide subjective context, in an attempt to ascertain if there is any (non-statistically representative) indication that program activities might be influencing the program results reported in this document. Any comparisons made in this report between non-participant and participant households that suggest that program outcomes might be attributable to program activities could be explored further in a future IE, or as part of a more robust evaluation design in the subsequent, follow-on program. 3. Endline Evaluation Findings Household Food Security and Vulnerability Status The overarching goal of Nobo Jibon is to reduce food insecurity and vulnerability in ten upazilas of the Barisal Division over five years. Critical to realizing this goal are improvements and increases in three areas: stunting in children 6-59 months, household food access, and household resilience, as measured by the CSI. Both baseline and endline surveys used anthropometric measures to assess the nutritional status of U5 children from sample households. This section reports the changes in those measures over the life of the program. Stunting rates improved in all districts over the program period. At baseline, all districts had high rates of overall stunting in children age 6-59: 38 percent in Barguna, 43 percent in Patuakhali, and as high as 50 percent in Barisal (Figure 2). Over the program period, overall stunting decreased from 44 percent to 35 percent across all sample households – a 20 percent overall reduction. This surpassed the program target of 40 percent and is comparable to national statistics – stunting fell nationally from 45 percent in 2010 to 35 percent in 2013.2021 20 FSNSP, 2014. 21 Confidence intervals for child stunting and all other program indicators are listed in Annex 2, as part of the IPTT table. If an indicator at endline measured across its entire confidence interval exceeds the program target, then for the purposes of this report, the indicator at endline is said to “exceed” the program target. If the program target falls within the confidence interval at endline, the indicator is said to have “met” the program target. 40 Figure 2: % of overall stunted (HAZ<‐2SD) children age 6‐59 months Note:  Stars indicate that difference between endline and baseline value is statistically significant at the 10% level. The improvement was slightly more marked in Barguna and Patuakhali, which both saw overall stunting decrease by 25 percent and 22 percent, respectively (Table 5). Severe stunting rates, which ranged from a low of 10 percent in Barguna to a high of 18 percent in Barisal at baseline, also saw substantial improvement at endline, with a 22 percent reduction in the overall sample, and greater improvements in Barisal (29 percent reduction) and Barguna (28 percent reduction; Table 5). Parallel to the baseline ranking, at endline Barisal remains the district with the highest rates of stunting (41 percent overall, 13 percent severe), and Barguna the lowest, with 28 percent stunted and 7 percent severely stunted (Table 5). For the overall sample, the severe stunting rate of 10 percent met the program target of 11 percent. Table 5: Program goal indicators, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baseline Endline % of overall stunted (HAZ<‐2SD) children age 6‐59 months 50.0 37.7 42.8 47.8 45.8 37.3 43.9 41.4 28.3 33.2 40.3 35.8 29.5 35.3 0 10 20 30 40 50 60 Barisal* Barguna*Patuakhali* Low* Middle* High* All* Baseline Endline 41 Table 5: Program goal indicators, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baseline Endline All households 43.9 35.3 ‐19.6 * 2,296 1,853 District         Barisal 50.0 41.4 ‐17.2 * 802 769      Barguna 37.7 28.3 ‐24.9 * 614 476      Patuakhali 42.8 33.2 ‐22.4 * 879 608 % of severely stunted (HAZ<‐3SD) children age 6‐59 months All households 12.9 10.0 ‐22.2 * 2,296 1,853 District         Barisal 17.7 12.6 ‐28.6 * 802 769      Barguna 9.8 7.1 ‐27.6 * 614 476      Patuakhali 10.7 9.1 ‐15.1 * 879 608 Household Food Insecurity Access Scale (HFIAS), mean value (0‐100) All households 28.7 19.4 ‐32.4 * 5,009 5,346 District         Barisal 26.2 16.4 ‐37.3 * 1,636 2,031      Barguna 36.6 20.1 ‐44.9 * 1,563 1,614      Patuakhali 24.2 22.3 ‐7.9 * 1,810 1,701 Coping Strategy Index (CSI), mean value (0‐100) All households 13.5 8.4 ‐37.8 * 4,969 5,346 District         Barisal 12.0 6.3 ‐47.5 * 1,623 2,031      Barguna 17.8 8.9 ‐50.1 * 1,561 1,614      Patuakhali 10.9 10.3 ‐5.6 1,785 1,701 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The Household Food Insecurity Access Scale (HFIAS) is reported on a scale of 0 to 100; higher scores indicate higher food insecurity, so a reduction in score is the desired outcome (see Annex 3 for details on the computation). The HFIAS value for the overall sample decreased by more than 30 percent, from 28.7 to 19.4 ( Figure 3). The HFIAS index at endline for all households sampled was 19 percent compared to a program target of 26 percent (Table 5). However, the magnitude of the endline-baseline difference varied substantially across regions, from a low 8 percent reduction in Patuakhali to markedly higher reductions in Barguna and Barisal (45 percent and 38 percent, respectively). 42 Figure 3: Household Food Insecurity Access Scale (HFIAS), mean value (0‐100) Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The Coping Strategies Index CSI scores were also scaled from 0-100, with a lower score indicating higher food security, hence lower scores are desirable. The pattern seen for CSI values mirrored that of the HFIAS: greater reductions in Barguna and Barisal (50 percent and 48 percent, respectively), and a small and statistically insignificant reduction in Patuakhali (Table 5). The CSI index for the overall sample at endline (8.4 percent) surpassed the program target of 12 percent (Figure 4). Figure 4: Coping Strategy Index (CSI), mean value (0‐100) 26.2 36.6 24.2 58.6 25.4 2.2 28.7 16.4 20.1 22.3 47.2 10.2 0.8 19.4 0 10 20 30 40 50 60 70 Barisal* Barguna*Patuakhali* Low* Middle* High* All* Baseline Endline 12.0 17.8 10.9 28.7 10.8 0.8 13.5 6.3 8.9 10.3 20.9 3.7 0.4 8.4 0 5 10 15 20 25 30 35 Barisal* Barguna* Patuakhali Low* Middle* High* All* Baseline Endline 43 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Another way to analyze achievement of program goals is by food security category (Table 6). Reductions in overall and severe stunting in children age 6-59 months were higher for medium and high food security categories. Medium food security households experienced the largest decrease in severe stunting from 14 percent to 10 percent, a 33 percent decrease. The only statistically significant decrease for stunting in the low food security category was for overall stunting, a 16 percent reduction, compared to more than 20 percent for the other two categories. This finding is logically consistent with the differences across food security categories for the scaled HFIAS and CSI scores. Both medium and high food security categories saw large improvements in HFIAS (60 percent and 62 percent decreases, respectively); the low food security category also saw an improvement but not as marked (a 19 percent change from baseline to endline). The changes in CSI across food security categories paint a similar picture, as households at all levels had significantly lower CSI scores, meaning they were turning to fewer coping strategies at endline than at baseline. The magnitude of change was substantial: from a 28 percent decrease in households with the lowest food security, to a 66 percent decrease for the medium group. Table 6: Program goal indicators, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baseline Endline % of overall stunted (HAZ<‐2SD) children age 6‐59 months All households 43.6 35.4 ‐18.9 * 2,213 1,848 Food security category          Low 47.8 40.3 ‐15.6 * 705 665       Medium 45.8 35.8 ‐21.9 * 782 582       High 37.3 29.5 ‐20.9 * 726 601 % of severely stunted (HAZ<‐3SD) children age 6‐59 months All households 12.6 10.0 ‐20.6 * 2,213 1,848 Food security category          Low 14.3 12.7 ‐11.5 705 665       Medium 14.4 9.6 ‐32.9 * 782 582       High 9.0 7.4 ‐17.8 * 726 601 Household Food Insecurity Access Scale (HFIAS), mean value (0‐100) All households 28.7 19.4 ‐32.4 * 4,944 5,336 Food security category          Low 58.6 47.2 ‐19.3 * 1,648 1,778       Medium 25.4 10.2 ‐59.9 * 1,648 1,779       High 2.2 0.8 ‐62.0 * 1,647 1,779 Coping Strategy Index (CSI), mean value (0‐100) 44 Table 6: Program goal indicators, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baseline Endline All households 13.4 8.4 ‐37.6 * 5,026 5,339 Food security category          Low 28.7 20.9 ‐27.2 * 1,648 1,779       Medium 10.8 3.7 ‐65.5 * 1,648 1,779       High 0.8 0.4 ‐48.7 * 1,647 1,777 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Analyzing the changes in HFIAS and CSI by gender (Table 7), we see that male-headed households have larger improvements than female-headed ones: the mean HFIAS decreased by 33 percent in male-headed households versus 23 percent in female-headed households; similarly, the mean CSI decreased by 39 percent (male-headed households) and 31 percent (female-headed households). Table 7: Program goal indicators, by sex of head of household   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline Household Food Insecurity Access Scale (HFIAS), mean value (0‐100) All households 28.7 19.4 ‐32.4 * 5,009 5,339 Sex head of household         Male 28.3 18.9 ‐33.2 * 4,705 5,001      Female 34.7 26.8 ‐22.9 * 304 339 Coping Strategy Index (CSI), mean value (0‐100) All households 13.5 8.4 ‐37.8 * 4,969 5,339 Sex head of household         Male 13.3 8.2 ‐38.4 * 4,666 5,001      Female 15.9 10.9 ‐31.4 * 303 339 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 45 Household Income and Expenditures This section reports data on household income and expenditures. Income and expenditure indicators disaggregated by district are presented below in Table 8. Additional tables presenting the indicators disaggregated by food security category (Table 72), and sex of household head (Table 73), and converted to U.S. dollars (Table 74) are made available in annex 6. Figure 5: Monthly Expenditure (in Tk) Per Capita (deflated) Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Overall and adjusted for inflation,22 the average monthly income per capita of sampled households increased by about 350 Tk (28 percent) over the life of the program (Table 8).23 The greatest relative gain across districts was in Barisal (1236 Tk to 2418 Tk, a 36 percent increase). Barguna and Patuakhali, meanwhile, improved per capita monthly income by 22 percent and 25 percent, respectively. Monthly expenditures per capita increased by about 200 Tk per month in the overall sample (Figure 5), however the increase was substantially greater in Barguna (27 percent) compared to Patuakhali (nine percent) and Barisal (seven percent). Barguna also exhibited the largest expenditures increase in absolute terms – about three times that of the other two districts. 22 A deflation factor was applied to the income and expense data based on inflation rates for 2009-2013 posted in the World Bank DataBank. The ideal span would be 2010-2014, however the rates for 2014 have not been posted as of this writing. 23 Income per capita is calculated based on the response to questions asking the annual value of income earned across a range of categories (17). The values across the categories are summed and divided by 12 to arrive at a monthly figure. This calculation is then divided by total household size to convert into a per capita value. Per capita expenditures are calculated similarly, summed across responses to questions regarding 9 categories of household expenses. Both of these calculations are consistent with how the indicators were calculated at baseline. 1520 1490 1546 1122 1368 2093 1528 1622 1895 1692 1425 1460 2308 1728 0 500 1000 1500 2000 2500 Barisal* Barguna*Patuahkali* Low* Medium* High* All* Baseline Endline 46 Table 8: Household income and expenditures (in Tk), by district   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline)    Number of observations (deflated)    Baseline Endline Monthly Income Per Capita24 All households 1274 2344 1628 27.8 * 5,026 5,338 District         Barisal 1236 2418 1679 35.9 * 1,649 2,019      Barguna 1247 2195 1524 22.3 * 1,565 1,610      Patuakhali 1332 2396 1664 24.9 * 1,812 1,709 Monthly Expenditures Per Capita All households 1520 2486 1727 13.6 * 5,026 5,338 District         Barisal 1520 2336 1622 6.7 * 1,649 2,019      Barguna 1490 2729 1895 27.2 * 1,565 1,610      Patuakhali 1546 2436 1692 9.4 * 1,812 1,709 Food Share (%) of Total Expenditures All households 62.3 54.2 ‐13.0 * 5,014 5,342 District         Barisal 63.6 58.6 ‐7.9 * 1,647 2,019      Barguna 62.4 50.2 ‐19.5 * 1,562 1,611      Patuakhali 60.9 52.6 ‐13.5 * 1,805 1,712 Asset Index25 All households 249.9 315.1 26.1 * 5,026 5,345 District         Barisal 307.3 368.0 19.7 * 1,649 2,019      Barguna 218.9 289.6 32.3 * 1,565 1,614      Patuakhali 224.4 276.9 23.4 * 1,812 1,712 Asset Index Per Capita All households 51.8 69.0 33.2 * 5,026 5,338 District         Barisal 60.2 74.9 24.4 * 1,649 2,019      Barguna 48.7 69.8 43.4 * 1,565 1,610      Patuakhali 46.8 61.3 30.9 * 1,812 1,709 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 24 All indicators reported in per capita terms, have been divided through by the number of reported persons in the household. 25 A description of how the asset index is calculated is available in Annex 3. 47 The data suggest that households across the sample directed their increased income in a manner consistent with program goals: at endline, food as a percentage of overall spending was lower for all districts by 13 percent on average. At baseline, at least 61 percent of expenditures in any district was for food; at endline this fell to as low as 50 percent, in Barguna. At the same time, Barguna had the greatest improvement in the per capita asset index: 43 percent versus 31 percent in Patuakhali and 24 percent in Barisal. These findings suggest that even though the increase in income in Barguna was the lowest of the three districts in terms of both absolute change and percentage change, compared to the other districts, in Barguna the increase had a stronger impact on households’ ability to direct a larger proportion of expenditures to investments in household assets. Similar gains were observed in all areas when disaggregated by food security group (Table 72). Increases in monthly income per capita ranged from 24 percent among the most food secure to 34 percent among medium food secure households. Monthly expenditures per capita were more varied: the least food secure households increased spending 27 percent, compared to increases of 10 percent among the most food secure and seven percent among medium food secure households. Notably, the households in the medium food security group were the only ones with monthly expenditures per capita that were lower than monthly income per capita.26 The least food secure households saw the largest decrease in food share as a percentage of total expenditures (16 percent) but contributed the largest share of expenditures to food (60 percent) compared to the most food secure households, which spent the least (46 percent). Likewise, the households in the low food security category saw the largest gains in asset index per capita (51 percent increase). This group, however, remained considerably lower in index value per capita than the most food secure households (43.5 index value per capita compared to 100.7 index value per capita, respectively). See Annex 3 for a description of the asset index computation. Considering the indicators by sex of household (Table 73), male-headed households generally saw more significant gains than female-headed-households. Monthly income per capita increased in male-headed households 29 percent, compared to 15 percent in female-headed households, though the latter reported higher income per capita (1746 Tk compared to 1620 Tk for male￾headed). Female-headed households saw monthly expenditures per capita increase 30 percent, while male-headed households’ expenditures grew 13 percent. Food share as percentage of total expenditures decreased 13 percent across all households – a desirable outcome –and the figure holds when analyzed by sex of household head (13 percent for both types). While the data show a 57 percent increase in asset index per capita in female-headed households (versus 32 percent for male-headed ones), this figure was not statistically significant. 26 Monthly expenditures per capita is used a proxy for income. 48 SO1 – Maternal and Child Health and Nutrition (MCHN) The MCHN component aims to contribute to improvements in antenatal care (ANC), infant feeding practices, and child healthcare related to immunization and treatment of diarrhea. This section reports the endline findings and compares them with the endline data, and analyzes the extent of changes in knowledge and practices in these health-seeking behaviors. Anthropometric Indicators The anthropometric data provide an indication of the combined impacts of SO1 and SO2 nutritional interventions and program activities. The baseline and endline surveys measured children under two years (U2) and under five years (U5) to assess the three standard indices of physical growth: weight for age (WAZ, or underweight), weight for height (WHZ, or wasting), and height for age (HAZ, or stunting). Stunting is a program goal-level indicator and is further disaggregated by age category for further discussion in this section (see Household Food Security and Vulnerability Status). Stunting, underweight and wasting are described below: Height for age (stunting): This index identifies whether a child has low height for her/his age. It is an important indicator of chronic malnutrition, and is a useful indicator in assessing changes in the magnitude of malnutrition over time. Weight for age (underweight): This index identifies whether a child is underweight for her/his age. It reflects both chronic and acute malnutrition, and is a useful indicator in assessing changes in the magnitude of malnutrition over time. However, it is not useful in distinguishing between stunting and wasting. (A child can be underweight for his/her age because he/she is stunted or wasted, or both stunted and wasted.) Weight for height (wasting): This index identifies whether a child has low weight for her/his height, and thereby helps identify children suffering from current or acute malnutrition or wasting. Weight for height is appropriate for examining short-term effects such as those from seasonal changes in food supply or short-term nutritional stresses brought about by illness. Table 9 reports, by food security category, the percentage of children in the 6-59 month, 6-23 month, 24-59 month age groups that are overall stunting (below -2 standard deviations from the median height for age per 2006 World Health Organization growth standards). In parallel fashion, Table 10 reports data for the same age groups by food security category, for those children with severe stunting (below -3 standard deviations from the median height for age) and severe wasting (below -3 standard deviations from the median weight for height). 49 The prevalence of overall stunting, for all children measured in the sample, decreased by 19 percent (Table 9). Improvements in overall child stunting, for the whole sample, are spread fairly uniformly in percentage terms across all of the food security categories, with children in the lowest food security category improving by 16 percent baseline to endline, children from the highest food security category improving 21 percent, and children in the middle food security category improving 22 percent. There were particularly notable declines in overall stunting achieved for children aged 6-23 months in the low and medium food security categories. Overall stunting declined 19 percent and 35 percent over the baseline for children from these two categories, respectively. The distribution of overall child stunting across food security category improved substantially for the 6-23 month age cohort, as the difference in stunting between the high and low food security category at baseline was 13 percentage points (38 percent vs. 25 percent) and decreased to 8 percentage points (31 percent vs. 23 percent). The particularly strong improvements seen in this age cohort could be a reflection of program effectiveness, as these children and their mothers might have had more time to benefit from SO1 programming, compared to older children represented in the data. For instance a mother with a child under 2 measured in the endline survey may have participated in SO1 programming from the time of her child’s conception, or perhaps before if she had another child under 5 in the household. Children measured closer to age 59 months might not have had the opportunity to benefit from the full range of SO1 programming, as the cumulative and irreversible effects of long-term malnutrition may have already affected them by the time them and their mother’s commenced program participation. The effectiveness of behavior change programming is likely more effective near the end of program, as compared to the initiation, as the program is scaling up. Table 9: Overall stunting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall stunting (HAZ<‐2SD) children age 6‐59 months All households 43.6 35.4 ‐18.9 * 2213 1848 Food security category          Low 47.8 40.3 ‐15.6 * 705 665       Medium 45.8 35.8 ‐21.9 * 782 582       High 37.3 29.5 ‐20.9 * 726 601 % of overall stunting (HAZ<‐2SD) children age 6‐23 months All households 33.0 25.7 ‐22.2 * 763 601 Food security category    50 Table 9: Overall stunting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       Low 37.9 30.8 ‐18.6 * 257 193       Medium 36.4 23.8 ‐34.7 * 248 193       High 24.9 22.7 ‐8.9 * 259 215 % of overall stunting (HAZ<‐2SD) children age 24‐59 months All households 49.2 40.1 ‐18.6 * 1,450 1,247 Food security category          Low 53.4 44.2 ‐17.2 * 448 471       Medium 50.1 41.7 ‐16.8 * 535 389       High 44.2 33.3 ‐24.6 * 467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*). Results disaggregated by child sex for child stunting, as well as, all of the other anthropometric indicators are available in Annex 6. Overall, there were no statistically significant differences for children under 5 between boys and girls for the anthropometric indicators, except for severe stunting – girls 6-59 months have a severe stunting prevalence of 9 percent while boys 6-59 months have a stunting prevalence of 11 percent. This difference was driven by a strong reduction (31 percent) in severe stunting for girls, from 13 percent at baseline to 9 percent (Table 48). The prevalence of severe stunting declined 20 percent for the entire sample, children aged 6-59 (Table 10). Particularly strong gains were realized in the medium food security category which experienced a 33 percent decline. When looking at severe stunting across different age categories, contrary to the findings for overall stunting, all of the improvement in severe stunting prevalence rates appears to be coming from children age 24-59 months, as compared to children 6-23 months. For children measured 24-59 months, severe stunting declined 24 percent, while there was no statistically significant change in child stunting rates for children 6-23 months. 51 Table 10: Severe stunting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of severe stunting (HAZ<‐3SD) children age 6‐59 months All households 12.6 10.0 ‐20.6 * 2,213 1,848 Food security category          Low 14.3 12.7 ‐11.5 705 665       Medium 14.4 9.6 ‐32.9 * 782 582       High 9.0 7.4 ‐17.8       726 601 % of severe stunting (HAZ<‐3SD) children age 6‐23 months All households 9.0 7.8 ‐13.1 763 601 Food security category          Low 11.8 12.0 1.3 257 193       Medium 9.3 6.5 ‐30.6 248 193       High 6.0 5.3 ‐11.2       259 215 % of severe stunting (HAZ<‐3SD) children age 24‐59 months All households 14.5 11.1 ‐23.7 * 1,450 1,247 Food security category          Low 15.8 13.0 ‐17.8 448 471       Medium 16.7 11.2 ‐32.9 * 535 389       High 10.8 8.6 ‐20.0 467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*). Table 11 reports, by food security category, the percentage of underweight children (below -2 standard deviations from the median weight for age per 2006 World Health Organization growth standards) across the 0-59 month, 0-23 month, and 24-59 month age groups. Similarly, Table 12 reports severe underweight (below -3 standard deviations from the median weight for age) for the same disaggregations of age group and food security category. 52 Figure 6: % overall underweight (WAZ<‐2SD) children age 0‐59 months Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The overall prevalence of underweight children in the total sample decreased from 39 percent to 27 percent (39 percent) over program life (Table 11: Overall underweight, by age and food security category). This compares favorably to national statistics, in which the prevalence of underweight children was unchanged at 32 percent from 2010 – 2013.27 The endline prevalence of overall underweight for children 0-59 months (27 percent) also compares favorably to the program target of 36 percent ( Figure 6). 27 FSNSP, 2013. 40.1 37.4 40.1 42.4 41.3 34.2 39.3 27.8 25.8 27.8 32.0 27.8 21.9 27.3 0 5 10 15 20 25 30 35 40 45 Baseline Endline 53 The range in percent overall underweight varied somewhat across food security categories. Children in the high food security category experienced the most improvement, with overall overweight prevalence declining 36 percent for this category. Similar to improvements seen in overall stunting rates, the cohort of children age 0-23 months appear to be driving much of the improvement for children in the overall sample. Improvements (declines in underweight prevalence) for children in this age category ranged from 36 percent for children in low food security households to 46 percent for children in high food security households. Table 11: Overall underweight, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline % of overall underweight (WAZ<‐2SD) children age 0‐59 months All households 39.4 27.4 ‐30.5 * 2,223 2,055 Food security category          Low 42.4 32.0 ‐24.6 * 707 722       Medium 41.3 27.8 ‐32.6 * 790 652       High 34.2 21.9 ‐36.0 *       727 680 % of overall underweight (WAZ<‐2SD) children age 0‐23 months All households 32.2 19.5 ‐39.4 * 770 807 Food security category          Low 38.1 24.3 ‐36.1 * 258 251       Medium 31.8 20.5 ‐35.5 * 253 262       High 26.7 14.5 ‐45.8 *    260 294 % of overall underweight (WAZ<‐2SD) children age 24‐59 months All households 43.2 32.4 ‐24.9 * 1,453 1,248 Food security category          Low 45.0 36.1 ‐19.7 * 449 471       Medium 45.8 32.8 ‐28.4 * 537 390       High 38.4 27.6 ‐28.2 *    467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*).   Moving to the “severe” level of this malnutrition indicator (Table 12), reductions in severe underweight prevalence were even more dramatic than for the overall prevalence of this indicator. Severe underweight prevalence fell 47 percent for the whole sample of 0-59 month children. Children in households across all categories of food security experienced significant declines, ranging from -32 percent for children in low food security households to -63 percent for children in medium food security households. 54 Table 12: Severe underweight, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline % of severe underweight (WAZ<‐3SD) children age 0‐59 months All households 9.8 5.2 ‐47.2 * 2,223 2,055 Food security category          Low 10.8 7.3 ‐32.5 * 707 722       Medium 11.7 4.3 ‐63.1 * 790 652       High 6.8 3.7 ‐45.5 *        727 680 % of severe underweight (WAZ<‐3SD) children age 0‐23 months All households 7.6 4.5 ‐39.9 * 770 807 Food security category          Low 8.9 8.1 ‐8.8 258 251       Medium 8.5 4.3 ‐49.2 * 253 262       High 5.3 1.7 ‐68.3 *        260 294 % of severe underweight (WAZ<‐3SD) children age 24‐59 months All households 11.0 5.6 ‐49.5 * 1,453 1,248 Food security category          Low 11.9 6.9 ‐42.4 * 449 471       Medium 13.2 4.3 ‐67.3 * 537 390       High 7.7 5.3 ‐31.4    467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*).   Reductions in overall wasting prevalence (Table 13) were pronounced, declining 32 percent for all children measured (6-59 months). Reductions in overall wasting also compare favorably to national statistics. At the national level, wasting prevalence increased from 10 percent to 12 percent from 2010-2011, and then remained flat at 12 percent from 2011-2013. At endline, reductions in overall child wasting (11 percent) exceeded the program target of 14 percent ( 55 Figure 7). The reduction was driven by children 24-59 months, for which wasting declined 42 percent over the life of the program. There were no statistically significant declines in wasting detected for measured children aged 6-23 months. Figure 7: % of overall wasting (WHZ<‐2SD) children age 6‐59 months Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 13: Overall wasting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall wasting (WHZ<‐2SD) children age 6‐59 months All households 16.2 11.0 ‐32.2 * 2,213 1,846 Food security category          Low 16.8 12.5 ‐25.4 * 705 663       Medium 18.3 10.1 ‐45.1 * 784 583 15.0 15.3 17.1 16.8 18.3 13.3 16.2 8.0 13.3 12.8 12.5 10.1 10.2 11.0 0 2 4 6 8 10 12 14 16 18 20 Baseline Endline 56 Table 13: Overall wasting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       High 13.3 10.2 ‐23.3 * 724 600 % of overall wasting (WHZ<‐2SD) children age 6‐23 months All households 15.4 13.8 ‐10.3 760 599 Food security category          Low 18.2 15.4 ‐15.3 256 192       Medium 16.5 14.2 ‐13.6 248 193       High 11.4 11.9 4.2 257 214 % of overall wasting (WHZ<‐2SD) children age 24‐59 months All households 16.6 9.7 ‐42.0 * 1,453 1,247 Food security category          Low 16.0 11.3 ‐29.1 * 449 470       Medium 19.2 8.0 ‐58.4 * 537 390       High 14.4 9.3 ‐35.4 * 467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*). With respect to severe wasting, there were no differences detected in the sample from baseline to endline except for in the medium food security category, for which severe wasting declined 67 percent (Table 14). While few differences were detected baseline to endline for this indicator, optimistically, levels of severe wasting are very low for the sample population, a little more than 1 percent for children aged 6-59 months. Notably, severe wasting was eliminated for children measured from households in the medium food security category for children aged 24-59 months. Table 14: Severe wasting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of severe wasting (WHZ<‐3SD) children age 6‐59 months All households 2.1 1.4 ‐32.4 2,213 1,846 57 Table 14: Severe wasting, by age and food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Food security category          Low 2.2 1.5 ‐30.7 705 663       Medium 2.7 0.9 ‐67.4 * 784 583       High 1.4 1.9 31        724 600 % of severe wasting (WHZ<‐3SD) children age 6‐23 months All households 3.1 3.0 ‐3.9 760 599 Food security category          Low 3.2 3.1 ‐2.8 256 192       Medium 4.5 2.7 ‐41.4 248 193       High 1.6 3.2 96.3        257 214 % of severe wasting (WHZ<‐3SD) children age 24‐59 months All households 1.6 0.7 ‐58.2 1,453 1,247 Food security category          Low 1.5 0.8 ‐46.8 449 470       Medium 1.9 0.0 ‐100 * 537 390       High 1.3 1.1 ‐13.7 467 387 Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*). Childhood Illness, Child Feeding Practices and Antenatal Care This section describes results of several indictors related to child and maternal health. A brief discussion on child illness measures is presented first, followed by several measures of child feeding and health of PLW. Table 15:  Incidence of child diarrhea, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of children under 5 with diarrhea in last 15 days All households 10.5 7.3 ‐30.8 *        2,312          2,186   Food security category    58 Table 15:  Incidence of child diarrhea, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       Low 12.7 8.6 ‐32.1 *            717              743         Medium 12.0 6.2 ‐48.3 *            822              694         High 7.0 7.0 0.3            772              749   % of afflicted children who sought treatment All households 72.9 73.0 0.1            244              160   Food security category          Low 74.0 71.6 ‐3.2              91                64         Medium 70.3 72.5 3.2              99                43         High 76.0 75.1 ‐1.2              54                52   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 16: Source of treatment for child diarrhea, by food security category Source of treatment (endline, %)   Indicator Low Middle High Total Pharmacy 46.6 44.8 32.5 41.4 Village doctor 22.4 49.0 22.8 29.6 MBBS doctor 10.1 2.9 26.6 13.7 Upazila health complex 6.2 2.9 5.6 5.1 Community clinic (CC) 6.7 3.4 4.5 5.1 Homopathic doctor 4.0 0.0 6.7 3.8 Clinic/hospital 1.8 0.0 8.4 3.5 Hospital/medical college 4.5 6.1 0.0 3.4 FWC 0.0 3.5 2.6 1.8 N 46 31 39 116 Diarrhea incidence decreased in U5s (-31 percent, Table 15), particularly in low and medium food security households (-32 percent & -48 percent, respectively). There were no significant changes in the proportion of households seeking treatment for diarrhea between baseline and endline, although in general most households did in fact seek treatment, nearly 75 percent. Pharmacy (41 percent), village doctor (30 percent), and MBBS doctor (14 percent) were the most frequent sources of treatment (Table 16). This same pattern of for treatment source is typical across all types of childhood illness reported in the survey 59 Table 17: Children with fever during the last two weeks, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of children under 5 with fever in last 15 days All households 54.8 48.8 ‐10.9 * 2,313 2,187 Food security category          Low 58.9 53.8 ‐8.6 * 717 743       Medium 56.6 45.9 ‐18.8 * 823 694       High 49.2 46.6 ‐5.3 772 750 % of afflicted children who sought treatment All households 65.0 75.7 16.5 * 1,467 1,068 Food security category          Low 60.8 68.5 12.7 * 485 400       Medium 65.5 79.7 21.7 * 533 319       High 68.9 80.2 16.4 449 349 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 18: Source of treatment for fever, by food security category Source of treatment (endline, %)   Indicator Low Middle High Total Pharmacy 41.4 40.0 38.4 39.9 Village doctor 32.6 27.3 25.3 28.4 MBBS doctor 6.2 8.3 16.4 10.4 Upazila health complex 6.4 6.5 6.5 6.5 Community clinic (CC) 4.8 4.9 4.8 4.8 Clinic/hospital 2.7 4.6 6.9 4.8 Homeopathic doctor 3.7 4.2 4.8 4.2 FWC 2.6 4.5 3.0 3.3 Other 1.1 2.7 1.0 1.6 Hospital/medical college 1.6 1.2 0.7 1.2 VHC (village health committee) 0.4 0.5 0.4 0.4 Satellite/EPI outreach centre 0.3 0.9 0.0 0.4 NGO static clinic 0.3 0.4 0.0 0.2 FWV 0.4 0.0 0.0 0.1 60 Table 18: Source of treatment for fever, by food security category Source of treatment (endline, %)   Indicator Low Middle High Total MCWC 0.0 0.4 0.0 0.1 NGO hospital 0.0 0.4 0.0 0.1 TBA trained 0.0 0.0 0.4 0.1 FWA 0.0 0.4 0.0 0.1 N 274 254 280 808 Likewise, households in the low and medium food security categories saw decreases in children with fever in the two weeks preceding the survey (Table 17). The percentage of children who sought treatment improved significantly for fever, increasing 17 percent over the life of the program. Treatment source for fever (Table 18) is similar to that for diarrhea, with pharmacy as the most prevalent (40 percent), followed by village doctor (28 percent) and MBBS doctor (10 percent). Table 19: Incidence of child cough/cold, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of children under 5 with cough/cold in last 15 days All households 54.8 56.3 2.7        2,313          2,187   Food security category          Low 58.9 60.9 3.4            717              743         Medium 56.6 51.8 ‐8.4 *            823              694         High 49.2 56.0 13.7 *            772              750   % of afflicted children who sought treatment All households 65.9 69.4 5.3 *        1,272          1,232   Food security category          Low 61.4 63.4 3.3            423              453         Medium 66.2 72.5 9.5 *            468              360         High 70.6 73.3 3.8            381              420   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 61 Table 20: Source of treatment for child cough/cold, by food security category Source of treatment (endline, %)   Indicator Low Middle High Total Pharmacy 42.2 43.2 39.3 41.5 Village doctor 31.0 26.3 25.4 27.5 MBBS doctor 5.7 8.0 15.4 9.9 Upazila health complex 6.1 7.5 7.4 7.0 Community clinic (CC) 5.6 6.0 4.0 5.1 Clinic/hospital 3.0 3.9 7.1 4.7 Homeopathic doctor 4.3 3.1 4.7 4.1 FWC 2.9 5.1 3.0 3.6 Hospital/medical college 2.1 0.4 1.0 1.2 Other 1.4 1.1 0.3 0.9 TBA trained 0.0 0.9 0.6 0.5 Satellite/EPI outreach centre 0.0 0.9 0.4 0.4 NGO hospital 0.0 0.4 0.4 0.3 MCWC 0.3 0.4 0.0 0.2 FWV 0.7 0.0 0.0 0.2 Neighbor 0.0 0.0 0.7 0.2 VHC (village health committee) 0.3 0.0 0.0 0.1 FWA 0.0 0.4 0.0 0.1 NGO static clinic 0.3 0.0 0.0 0.1 N 46 31 39 116 Cough/cold among U5 children is more mixed, with a significant increase among the most food secure households ( 14 percent, Table 19) and decrease among medium food secure households ( -8 percent). However, for the overall sample there was no change in prevalence of cough/cold. Treatment source for cough cold ( 62 Table 20) mirrored that of diarrhea and fever, with pharmacy (42 percent), village doctor (28 percent), and MBBS doctor (10 percent) cited as the most frequent sources. Table 21 displays information on exclusive breastfeeding of children under six months. There was no statistically significant change in exclusive breastfeeding for the total sample population from baseline to endline. Medium food security households showed increases in breastfeeding practices from baseline to endline, with a 38 percent increase in exclusive breastfeeding under six months. At endline, households in the medium category were also the most likely of all food security categories to exclusively breastfeed (52 percent). Meanwhile, the least food secure households were least likely (39 percent) to breastfeed children exclusively. Table 21: Breastfeeding practices, by food security category28   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline Children under 6 month exclusively breastfed All households 38.6 44.9 16.5 276 320 Food security category          Low 40.4 39.1 ‐3.2 75 104       Medium 37.7 52.0 37.6 * 104 94       High 38.0 44.5 17.1 97 122 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Data on three measures of child feeding and care giving practices are shown in Table 22. First, infants and toddlers six-to-23-months-old who receive a minimally acceptable diet (apart from breast milk); second, infants and toddlers older than six months who received iron rich/ iron fortified foods during the previous day; and third, households consuming adequately iodized salt. Significant increases were seen among all households overall in every category. Figure 8: Infants/toddlers 6‐23 months who receive a minimally acceptable diet 28 At the time of design of the project, the indicator for EBF was defined by FFP to be for children 0‐6 months. This definition was subsequently changed to be for children under 6 months (0‐5 months). 63 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Households at every food security level showed large increases in the percentage of children who received a minimally acceptable diet (Table 22). The percentage of children receiving a minimally acceptable diet in least food secure household grew from two percent to 16 percent, a 920 percent increase, while the percentage of medium and high food security households tripled or more to 17 percent and 32 percent, respectively (Table 22). Two-thirds of all households had infants older than six months who received iron rich/iron fortified foods in the previous day. Households in both the low and medium categories saw increases of about one-third (31 percent and 39 percent, respectively. Table 22: Child feeding and care giving practices, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline Infants/toddlers 6‐23  months who receive a minimally acceptable diet (apart from breast milk) All households 5.8 22.5 290.1 * 784 687 Food security category          Low 1.6 16.1 920.2 * 261 209       Medium 5.0 17.2 246.4 * 261 214       High 10.7 31.9 197.2 * 263 264 Infants/toddlers older than 6 months who received iron rich/iron fortified foods during the previous day   All households 52.1 64.6 24.1 * 784 677 Food security category          Low 44.8 58.9 31.4 * 261 207 6.1 5.4 5.9 1.6 5.0 10.7 5.8 20.6 21.3 26.5 16.1 17.2 31.9 22.5 0 5 10 15 20 25 30 35 Baseline Endline 64       Medium 48.9 67.9 38.8 * 261 212       High 62.5 66.6 6.6 263 258 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 23 shows several indicators used to measure nutrient consumption of pregnant and lactating women (PLW). All households and all food security categories experienced substantial increases in consumption of food rich in iron, consumption of food rich in vitamin A, and use of iron or iron folate supplements in the last seven days. Figure 9: %age of PLW that consume food rich in iron Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The most marked changes came in the consumption of foods rich in iron and consumption of foods rich in vitamin A. In the former category, the overall increase of 184 percent reflects a change from 32 percent at baseline to 91 percent at endline (Table 23). The least food secure households were the least likely to consume iron-rich foods (85 percent). In the latter category, substantial increases include the overall change from 22 percent of households to 60 percent (166 27.4 24.3 41.6 24.0 35.6 34.7 31.9 91.3 90.5 89.4 85.3 92.5 93.5 90.5 0 10 20 30 40 50 60 70 80 90 100 Baseline Endline 65 percent increase). Again, low food security households showed the largest change increasing by 227 percent to almost half of households. At endline, 91 percent of PLW surveyed reported consuming food rich in iron (Figure 9), substantially exceeding the program target of 60 percent. Vitamin A supplementation among mothers of U2s increased overall by 45 percent (Table 23). The most food secure households experienced a 55 percent increase, compared to 47 percent for the medium food security households, and 29 percent for the least food secure households. Significant increases in use of iron or iron folate supplements is also shown in Table 23, though the overall prevalence was just 12 percent of households. Table 23: Nutrient consumption among PLW, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Percentage of PLW who: Consume food rich in iron All households 31.9 90.5 183.9 * 420 517 Food security category          Low 24.0 85.3 254.8 * 124 168       Medium 35.6 92.5 160.1 * 162 165       High 34.7 93.5 169.8 * 134 184 Consume food rich in vitamin A All households 22.4 59.6 165.7 * 420 517 Food security category          Low 14.9 48.9 227.4 * 124 168       Medium 22.0 55.8 153.2 * 162 165       High 29.8 72.8 144.0 * 134 184 Consume food rich in calcium All households 12.3 12.4 1.1 420 517 Food security category          Low 8.5 6.1 ‐27.8 124 168       Medium 10.2 5.9 ‐42.8 162 165       High 18.3 24.1 31.7 134 184 Have taken iron or iron folate supplements in the last 7 days All households 2.2 11.8 448.6 * 420 517 Food security category          Low 0.8 8.6 1041.5 * 124 168       Medium 1.8 10.7 497.0 * 162 165       High 3.9 15.7 305.0 * 134 184 % of mothers of children aged 6‐23 months who received high‐dose Vitamin A supplement within 8 weeks postpartum (6 weeks if not exclusively breastfeeding) in last pregnancy All households 26.3 38.2 45.4 * 696 710 66 Table 23: Nutrient consumption among PLW, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Food security category          Low 26.5 34.2 29.1 * 236 225       Medium 23.7 34.8 47.0 * 229 220       High 28.7 44.5 55.0 * 230 265 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Recognizing the importance of adequate antenatal care (ANC) to health and well-being of both infants and mothers, Nobo Jibon sought to support greater access to appropriate medical care among PLW. Table 24 shows the percent of pregnant women or mothers of children under two￾years-old who attended at least four antenatal care sessions. Overall, one-third of respondents reported attending ANC sessions at endline, a 176 percent increase from baseline. Each food security category exhibited similar results, with slightly more households in the high category attending (35 percent) than households in the medium and low categories (33 percent and 30 percent, respectively). Compared to baseline, this represents double the percentage of households among the most food secure households and more than triple that among the other two categories. Table 24: Attendance at antenatal care sessions, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % of pregnant women or mothers of children under 2 attending at least 4 ANC sessions All households 11.9 32.9 175.8 * 1,125 1,093 Food security category          Low 8.5 30.1 256.4 * 365 336       Medium 9.8 32.8 233.5 * 397 349       High 17.7 35.3 99.0 * 362 409 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Vitamin A supplementation and deworming services are also part of Nobo Jibon’s plan to improve diet and reduce illness. While the increase in the percentage of children who received Vitamin A supplementation among all households was minimal (Table 25), both low food security and medium food security households experience significant changes. Notably, these changes were significant in different directions: the least food secure households experience a 22 percent decrease in Vitamin A supplementation, and households in the medium category saw a 36 percent increase. All households saw positive increases in the percent of children 12-23 67 months-old who received deworming within the last six months. Across the entire sample, one￾third of children in the age group received deworming, a 74 percent increase. The most food secure households increased the most (97 percent) compared to the other food security categories, and those households were most likely to have dewormed children (41 percent). Table 25: Percentage of children 12‐23 months who received Vitamin A supplementation, deworming treatment within last 6 months, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of children that received Vitamin‐A supplementation All households 43.4 45.4 4.6 513 415 Food security category          Low 47.3 37.0 ‐21.8 * 168 127       Medium 37.7 51.4 36.1 * 162 133       High 44.8 47.1 5.2 183 155 % of children 12‐23 months who received deworming w/in last 6 months All households 18.9 32.8 73.8 * 513 419 Food security category          Low 16.7 24.0 44.1 166 127       Medium 18.8 31.3 66.5 * 162 133       High 20.9 41.0 96.6 * 186 159 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Water, Sanitation and Hygiene (WASH) The following tables provide information on WASH indicators including hygiene, latrines, and quality of drinking water. Table 26: Caregiver hygiene practices, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of caregivers demonstrating proper personal hygiene behaviors All households 31.3 38.1 21.8 * 2,341 2,140 Food security category          Low 21.9 27.6 25.9 * 729 727       Medium 32.6 36.9 13.1 * 831 685       High 38.5 49.7 29.0 * 782 729 % of caregivers demonstrating proper food hygiene behaviors All households 20.4 26.6 30.4 * 2,341 2,054 Food security category    68 Table 26: Caregiver hygiene practices, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       Low 16.4 18.8 14.2 729 686       Medium 20.3 24.3 20.0 * 831 657       High 24.1 36.1 49.8 * 782 711 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The caregiver hygiene practices measured and shown in Table 26 include personal hygiene behaviors, food hygiene behaviors, water hygiene behaviors, and environmental hygiene behaviors. Across all these practices, surveyed households showed statistically significant increases from baseline to endline. Caregivers in low food security households were least likely to use each practice, though the percentage increased significantly for all behaviors except food hygiene. Likewise, caregivers in high food security households were most likely to demonstrate practices in all but one of the behaviors. Personal hygiene behaviors increased significantly (Table 26), but by the least overall of the four measured behaviors (22 percent). The largest increase in this area was among high food security households, from 39 percent to 50 percent, a 29 percent increase. Low food security households followed, with a 26 percent increase in households demonstrating the behavior. Food hygiene behaviors increased more than 30 percent among all surveyed households. Again, high food security households showed the largest increase from 24 percent to 36 percent, a 50 percent increase. Relative to baseline, a greater percentage of medium food security and low food security households also adopted the behavior (20 percent increase and 14 percent increase, respectively), though the latter improvement was not statistically significant. Qualitative Information (SO1) Pregnant and lactating women (PLW), husbands of PLW, and adolescents all reported positive experiences from Nobo Jibon activities. Through the program, PLW met monthly to learn and discuss several topics related to pregnancy, childbirth, and child feeding, as well as hygiene and immunization. Husbands were invited to attend these courtyard meetings, but they were not obligated; generally, every husband attended at least part of one session, though few husbands were regular attendees. Both men and women found the meetings valuable, reporting understanding of several topics related to maternal and child health. Growth monitoring and promotion (GMP), in particular, was well received. The training and counseling sessions led to changes in the approach of husband, including greater awareness of their role as father and responsibility to support their wife. Husbands also reported greater awareness of the importance of hygiene and sanitation around the home. 69 The greatest barriers to behavior change among PLW included the distance and cost of clinic visits and also occasional lack of family support. Husbands reported poverty and economic insolvency as obstacles to changing practices, despite valuing the training sessions and trying to support their wives. Men and women recommended continuing counseling sessions and food rations. Involvement of the village development committees (VDC) and VHC is also important. According to PLW, sustainability depends on motivating involvement in VDCs and/or VHCs without food rations as an incentive. Another factor influencing the sustainability of improved health and hygiene practices, according to PLW, include whether a linkage can be built between VHCs and public programs, namely the National Nutrition Network. The continuation of GMP sessions, which are an important means of educating and involving husbands in the promotion of proper health and hygiene behaviors in the household, was also cited as an important form of sustaining improvements achieved by SO1 programming. Promotion of MCHN and gender learning also included adolescents who ranged in age from 13 years old to post-secondary school age. Groups met outside of regular school hours and discussed a wide range of topics including community and the environment, personal hygiene, water and sanitation, health and nutrition, and gender inequality. This final category included issues such as educational inequality, dowry, early marriage, violence against women, and mobility outside the home. Both boys and girls believed that the information discussed was important but primarily of value to girls, as “direct beneficiaries of change.” All group members mentioned improved awareness of social gender-related issues and greater confidence in addressing such topics in the community. Lack of cooperation from parents and community leaders, especially regarding gender issues, was the biggest obstacle to change among youth. The focus group discussion (FGD) participants expressed that teachers often supported the student’s participation in the groups, however parents not so much, some fearing that it would take away time from studies. Adolescents recommended continuing group activities, suggesting that it would be worthwhile to educate adults with respect to the value of the adolescent groups, so that the groups could gain wider acceptance in the community, as well as, establish more formal linkages to the VDC, VHC, and VDMC. SO2 – Market-based Production and Income Generation SO2 seeks to enhance household productivity and income in order to improve food access for poor households. Performance measures include those defined for each Intermediate Result and comprehensive indicators to estimate market-based production and income generation: number of income sources per household, annual income from the sale of agricultural products, Household Dietary Diversity Score (HDDS), and months of adequate household food provisions (MAHFP). The next two tables report data disaggregated first by food security category (Table 27), then by sex of household head (Table 28). 70 Notable gains were seen over program life in the value of agricultural sales, especially for low and middle food security terciles, where sales value increased, in real terms, by almost 1/3 (from average 3942 Tk to 5089 Tk in the lowest tercile and from 7410 Tk to 9871 Tk in the middle tercile).29 While male-headed households reported significant gains (10,808 Tk to 12,139 Tk), agricultural product sales in female-headed households were relatively unchanged (the -9.8% difference between baseline and endline is not statistically significant). As in the baseline, the agricultural sales income of the tercile of households in the highest food security category was substantially higher than that of households in the lowest category: at baseline, average agricultural income of the low food security tercile was just 19 percent of that of the high food security tercile (3952 Tk compared to 20216 Tk), and at endline, this gap had narrowed only slightly, to 25 percent (5089 Tk compared to 20098 Tk). Figure 10: Household Dietary Diversity Score (HDDS) Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 29 Consistent with reporting in the baseline, value of agricultural sale products here includes sales of both crops and livestock. The indicator reported in the IPTT includes only sales of crops. 5.0 4.5 4.7 3.8 4.5 5.8 4.7 6.1 5.3 5.5 4.7 5.4 7.0 5.7 0 1 2 3 4 5 6 7 8 Baseline Endline 71 Dietary diversity, as measured by the HDDS, saw positive change, its average value increasing by one unit or nearly one (on a scale of 0-12) over program life (Table 27). The minor exception is in the high food security category, where the average HDDS performed even better, increasing from 5.8 to 7.0. As expected, the baseline and endline data both show that the more food￾insecure the household, the lower the HDDS. The average overall HDDS is 5.7 ( Figure 10). As in the baseline, there was little difference in average HDDS of men (5.7) and women (5.4) at endline. Over program life, the average number of months of food provisioning increased by about one month for households with low and middle food security, and stayed about the same for those with higher food security. While HDDS at endline (5.7) exceeded the program target (5.5), MAHFP (10.4 months) did not (11 months). Table 27: Economic and food access indicators, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Average value of agricultural product  sales (Taka) All households 10521 11646 10.7 * 4,944 5,333 Food security category          Low 3942 5089 29.1 * 1,648 1,785       Medium 7410 9871 33.2 * 1,648 1,787       High 20216 20098 ‐0.6 1,647 1,760 Household Dietary Diversity Score (HDDS) All households 4.7 5.7 20.7 * 4,944 5,336 Food security category          Low 3.8 4.7 24.2 * 1,648 1,778       Medium 4.5 5.4 18.0 * 1,648 1,779       High 5.8 7.0 20.4 * 1,647 1,779 Months of Adequate Household Food Provisions (MAHFP) All households 9.4 10.4 10.2 * 4,944 5,336 Food security category          Low 7.1 8.4 17.9 * 1,648 1,778       Medium 9.6 10.9 13.8 * 1,648 1,779 72 Table 27: Economic and food access indicators, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline       High 11.6 11.9 2.5 * 1,647 1,779 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*). The value of agricultural product sales are reported as deflated, real values. Table 28: Economic and food access indicators, by sex of head of household   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline Average value of agricultural product  sales (Taka) All households 10448 11642 11.4 * 5,026 5,334 Sex head of household         Male 10808 12139 12.3 * 4,722 4,993      Female 4850 4377 ‐9.8 304 342 Household Dietary Diversity Score (HDDS) All households 4.7 5.7 20.8 * 5,026 5,339 Sex head of household         Male 4.7 5.7 20.7 * 4,722 5,001      Female 4.4 5.4 22.7 * 304 339 Months of Adequate Household Food Provisions (MAHFP) All households 9.4 10.4 10.2 * 5,026 5,339 Sex head of household         Male 9.5 10.4 10.2 * 4,722 5,001      Female 8.6 9.5 10.3 * 304 339 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*).The value of agricultural product sales are reported as deflated, real values. Agricultural Production and Marketing Practices The endline QPE included a range of questions related to knowledge of agricultural production and marketing practices, access to quality inputs, capital, and markets; access to natural resources and/or productive assets; and – as a proxy indicator of improved household productivity and income – questions about dietary diversity. Results are presented in this section. The evaluation sought information about use of improved agricultural techniques. Similar to baseline measurements, few households use three or more improved agricultural practices (Table 73 29). While the overall increase among sampled households was significant (42 percent), this reflects seven percent of all households, with little variation between food security categories (Figure 11). Figure 11:  % of HH adopting 3 or more improved practices Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 30 shows that, by far, traditional agricultural techniques such as fertilizer and chemical pest control were the most common (81 percent and 68 percent, respectively). Changes in these techniques, while significant, were small. Changes in usage of other techniques varied in magnitude and direction. Composting and animal manure were both used by about one-third of households (33 percent and 30 percent, respectively), though composting decreased 11 percent while animal manure increase four percent. Biological pest control and crop rotation were used by 11 percent and 10 percent of households, respectively, and both saw marked increases (70 percent and 160 percent, respectively). Table 29: Use of improved agricultural techniques, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH adopting 3 or more improved practices All households 4.9 6.9 42.2 * 2,011 3,065 Food security category          Low 5.9 6.8 14.9 394 806 5.6 5.7 3.5 5.9 3.8 5.1 4.9 11.3 3.8 5.8 6.8 5.8 7.9 6.9 0 2 4 6 8 10 12 Baseline Endline 74       Medium 3.8 5.8 51.1 * 661 1,027       High 5.1 7.9 54.0 * 956 1,232 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 30: Type of improved agricultural technique used % household reporting using technique (endline)   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Fertilizer 83.2 81.2 ‐2.4 * Chemical pest control 65.1 67.8 4.1 * Compost 36.5 32.5 ‐11.0 * Animal manure 28.5 29.6 3.9 * Biological pest control 6.2 10.5 68.8 * Crop rotation 3.9 10.3 161.4 * Integrated pest management 12.4 10.1 ‐18.8 * Mechanical pest control 1.9 3.5 86.1 * Improved irrigation 3.0 2.5 ‐17.3 N 2011 3065 Among all sample households, thirteen percent reported they have received any agricultural training (Figure 12). Less than one-quarter of households (22 percent) engaged in agricultural production in previous year received training. Of those households, the most common source of training by far was Nobo Jibon (62 percent). This is also true among each food security group, with greater popularity among the least food secure group (69 percent) than the medium and high groups (58 percent and 59 percent, respectively). Government training was next most popular overall, though medium (35 percent) and high (33 percent) food security households took advantage of this source more than low food security households (21 percent). After government training (30 percent), NGOs (16 percent) and seed companies (7 percent) were the next most popular sources of training overall and among each food security group. 75 Figure 12: Source of agricultural training reported by households at endline, by food security category N= household reporting receipt of agricultural training (lowest = 176, middle = 210, highest = 182). Households were far more likely to sell agricultural produce to a local market (79 percent) than any other option (Table 31), equal to baseline measurements. More than one-quarter of households also sold to traders (28 percent) or to neighbors or relatives (27 percent). Sales to either an itinerant buyer or to NGOs, cooperatives or sales companies accounted for less than three percent of households. Table 31: Types of buyers for agricultural product % household reporting using buyers (endline)   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Local market 78.6 78.6 0.0 Traders 23.9 28.3 18.5 * Neighbors/relatives 18.2 27.2 49.6 * Local broker 8.3 8.3 ‐0.1 Itinerant buyer 1.5 2.5 67.9 * Other (NGO, collection point, sales company) 0.5 2.2 307.7 * N 1177 1819 Calculated as a percentage of households reporting agricultural sales, both Nobo Jibon beneficiaries and non‐beneficiaries 14.7 14.3 18.7 16.2 6.0 7.8 7.4 7.2 69.3 58.0 59.2 61.5 20.7 34.5 33.0 30.2 0 10 20 30 40 50 60 70 80 Lowest Middle Highest Total Sample NGO Seed Company Nobo Jibon GOB 76 Table 32 shows changes in marketing practice by food security category. Significant gains were reported in households adopting improved marketing practices, though these are both still remarkably small proportions of the population, with just two percent of all households reporting either measure. Most notably, both low and medium food security households reported some involvement in this activity, after exhibiting no involvement at baseline. Table 32: Use of marketing practices, by food security category30   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH adopting improved marketing practices All households 0.4 1.9 324.7 *   1,177 1,821 Food security category                Low 0.0 2.1 N/A    196 386       Medium 0.0 2.3 N/A    352 593       High 0.8 1.5 77.3 *   629 843 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 30 This indicator definition was revised after the baseline survey. It is calculated and reported here based on the old definition to maintain consistency with the baseline survey. The results change only marginally when calculating as a beneficiary-based measure and do not change the interpretation of results. 77 Table 33 shows the source of agricultural inputs for program households. Local markets (88 percent) and neighbors or relatives (10 percent) were the most popular sources, in both baseline and endline rounds . Reliance on local markets increased from 79 percent to 88 percent, while sourcing inputs from neighbors or relatives decreased by 40 percent, from 17 percent to 10 percent. Use of GOB sources has declined from 8 percent at baseline to 5 percent at endline. Other sources, including NGOs, cooperatives and farmer groups, itinerant merchants, companies, itinerant merchants, and VDC were not widely accessed either at baseline or endline. Table 33: Source of agricultural inputs % household reporting purchase or receipt of inputs (endline)   Indicator Baseline Endline Percent difference (Endline ‐ Baseline)      Local market 79.4 88.0 10.8 * Neighbor/relatives/individuals 17.3 10.4 ‐39.6 * Trained input retailers 7.1 8.4 18.4 Nobo Jibon 0.0 5.7 100.0 * GOB 7.9 4.7 ‐39.9 * NGOs 2.4 2.1 ‐15.2 Itinerant merchants 0.5 1.8 274.1 * Other 4.8 0.7 ‐85.7 * Companies 0.2 0.6 135.6 Village development committees 0.2 0.5 108.6 Cooperative/farmer group 1.8 0.3 ‐82.9 * N 1613 2507    Table 34 provides general information about household engagement in agricultural activities. While more than two-thirds of households have agricultural land, access varies widely across food security categories. All groups saw a significant increase from baseline, but the most food secure households were most likely to have land (80 percent) compared to the households in the middle and low food secure categories (68 percent and 56 percent, respectively). A similar trend follows for average land area (Table 34). The least food secure households reported 69.8 decimals of land (a 59 percent increase), while the medium and high food security households had 89 decimals and 129 decimals, respectively. These trends also continue with average value of agricultural product sales. Overall, the average value rose 60 percent. Low and medium food security households experienced the largest improvements (86 percent and 92 percent, respectively). At 29,000 taka (43 percent increase), the most food secure households’ average sales was double that of medium food security households and nearly four times greater than the least food secure households. 78 Use of khas land and water bodies for agricultural production dropped significantly overall (25 percent) and among each food security group (Table 34). High food secure households were least likely to use khas land and water bodies (38 percent) and also reported the largest decrease of any group (34 percent). Use among medium food secure households also decreased nearly one-third (31 percent). The least food secure households reported the greatest percentage of households using khas land and water bodies (55 percent), despite also experience a nine percent decrease. Table 34: Summary statistics for agriculture, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH with agricultural land All households 59.3 68.0 14.5 * 4,943 5,336 Food security category          Low 47.7 55.7 16.8 * 1,648 1,778       Medium 55.6 68.1 22.6 * 1,647 1,779       High 74.8 80.1 7.0 * 1,647 1,779 Average land area (decimals) All households 88.1 99.5 12.9 * 2,930 3,610 Food security category          Low 43.9 69.8 59.0 * 783 982       Medium 70.2 88.9 26.6 * 915 1,208       High 129.6 129.1 ‐0.4 1,232 1,419 Average value of agricultural product  sales (Tk) All households 10,521 11,646 10.7 * 4,944 16,770 Food security category          Low 3,942 5,089 29.1 * 1,648 1,785       Medium 7,410 9,871 33.2 * 1,648 1,787       High 20,216 20,098 ‐0.6 1,647 1,760 % of households using khas land/water bodies for production of crops, livestock, and fish All households 61.8 46.7 ‐24.5 * 4,944 5,336 Food security category          Low 59.7 54.5 ‐8.7 * 1,648 1,778       Medium 68.6 47.6 ‐30.5 * 1,648 1,779       High 57.3 38.0 ‐33.7 * 1,647 1,779 79 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*). Agricultural product sales are reported as deflated, real values. As can be seen in Table 35, fishing is the most common form of community property use in the program area. Across all households in the program area, 46 percent of households surveyed at endline use water bodies for fishing. This is down from 61 percent of households at baseline and is the principal driver of the reduction in the indicator in Table 34 measuring the proportion of households using khash land/water bodies for production of crops, livestock, and fish. It should be noted that the endline survey included a response category “do not use” that was not included for this particular question (“What are water bodies used for?”) in the baseline survey. At endline, 43 percent of respondents indicated they do not use their water bodies, a response that was unavailable for those surveyed at baseline. Table 35: Khash land/water body use, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of households using khas land/water bodies for gardening All households 0.9 1.9 118.6 * 4,944 5,336 Food security category          Low 0.4 3.1 630.7 * 1,648 1,778       Medium 1.2 1.4 15.7 1,648 1,779       High 1 1.2 24.1 1,647 1,779 % of households using khas land/water bodies for ag production All households 1.8 2.2 23.3 4,944 5,336 Food security category          Low 0.9 2.3 163.6 * 1,648 1,778       Medium 2.3 1.6 ‐31.4 1,648 1,779       High 2.2 2.7 25.6 1,647 1,779 % of households using khas land/water bodies for fishing All households 61.2 45.8 ‐25.1 * 4,944 5,336 Food security category          Low 59.3 53.6 ‐9.6 * 1,648 1,778       Medium 67.7 46.7 ‐31 * 1,648 1,779       High 56.5 36.9 ‐34.7 * 1,647 1,779 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 80 Household agricultural production is shown in Table 36, disaggregated by food security category. Significant increases were observed overall and in each individual category shown, among all three food security groups. Nearly 90 percent of all households (89 percent) engaged in some form of production (crops, livestock or fish), an increase of 17 percent from baseline (Table 36). The largest increase was among the least food secure households (25 percent), with 86 percent reporting production. In addition, 44 percent of households reported an increase in production, up from 39 percent at baseline. This included more than half (53 percent) of the most food secure households, 43 percent of medium food security households, and 35 percent of the least food secure households that experienced increases in production. Similar to baseline, livestock production was the most common area, with 83 percent of households engaging, an increase of more than one-third (34 percent), and little variation among food security groups. The next most common form of production was crops (57 percent). Of those households engaging in livestock production, those that reported increases fell to 22 percent from 27 percent at baseline. However, this statistic may be misleading, since it is reported as a percentage of households engaged in livestock production, rather than all sampled households. Due to the large increase in households engaging in livestock production (83 percent vs. 62 percent at baseline, Table 36), the absolute number of households reporting increases in livestock production certainly grew even if the proportion of those engaged in livestock production declined marginally. The distribution of household production of crops across food security category improved dramatically, due to an impressive growth rate in the proportion of low food security households engaging in agricultural production from baseline to endline (Table 36). At baseline, less than a quarter (24 percent) of low food security households were engaged in agriculture, compared to 58 percent of high food security households – a difference of 34 percentage points. By endline, nearly half (45 percent) of the least food secure households produced crops, as this group experienced the largest growth of any group (increase of almost 90 percent). Meanwhile, crops were produced by 69 percent of high food security households at endline. However, the growth in households engaged in agricultural production in this highest food security category was the lowest at 19 percent. The resulting inequality in agricultural production between high food security households and low food security households lessened from 34 percent at baseline to 24 percent (69 percent – 45 percent) at endline Fish production was the least popular form of agriculture (30 percent), but still increased 34 percent (Table 36). All groups saw significant increases in fish production, with a notable 84 percent increase among the least food secure households. Further, nearly one-quarter of all households (23 percent) reported an increase in fish production, compared to 15 percent at baseline, a 48 percent increase. 81 Table 36: Household production, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH with agricultural production last year All households 40.7 57.4 41.2 * 4,944 5,336 Food security category          Low 23.9 45.3 89.6 * 1,648 1,778       Medium 40.1 57.7 44.0 * 1,648 1,779       High 58.0 69.3 19.3 * 1,647 1,779 % reporting increased agricultural production All households 40.0 47.9 19.7 * 2,011 3,065 Food security category          Low 36.5 40.4 10.6 * 394 806       Medium 38.1 46.6 22.5 * 661 1,027       High 42.8 53.9 25.9 * 956 1,232 % HH with livestock All households 61.7 82.5 33.8 * 4,944 5,336 Food security category          Low 60.7 80.7 32.9 * 1,648 1,778       Medium 60.1 82.7 37.6 * 1,648 1,779       High 64.1 84.1 31.2 * 1,647 1,779 % reporting increased livestock production All households 27.1 21.8 ‐19.5 * 3,048 4,403 Food security category          Low 23.1 19.3 ‐16.6 * 1,001 1,435       Medium 28.8 21.5 ‐25.4 * 991 1,472       High 29.4 24.7 ‐16.1 * 1,056 1,496 % HH with fish production All households 22.8 30.4 33.5 * 4,944 5,336 Food security category          Low 10.6 19.5 84.4 * 1,648 1,778       Medium 20.1 28.1 40.1 * 1,648 1,779       High 37.8 43.6 15.6 * 1,647 1,779 % reporting increased fish production   All households 15.3 22.6 47.7 * 1,127 1,624 Food security category          Low 17.1 18.8 9.9 * 174 347       Medium 14.1 21.3 51.4 * 331 500       High 15.5 25.2 62.7 * 622 776 % HH engaged in at least one category (crops, livestock, fish) All households 75.9 88.6 16.7 * 4,944 5,336 82 Table 36: Household production, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Food security category          Low 68.6 85.8 25.0 * 1,648 1,778       Medium 74.5 88.8 19.2 * 1,648 1,779       High 84.6 91.2 7.9 * 1,647 1,779 % reporting increased production in any category All households 39.1 43.6 11.4 * 3,752 4,728 Food security category          Low 31.0 34.8 12.3 * 1,131 1,525       Medium 39.1 42.7 9.1 * 1,228 1,580       High 45.6 52.7 15.4 * 1,393 1,623 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Qualitative Information (SO2) Program population was divided into three groups for this objective: 1) extreme poor (EP), who had limited experience in agriculture and access to resources and were all women; 2) homestead production poor (HPP), comprised of all women; and 3) productive poor (PP), the majority of who were men, around 60%, and also larger-scale farmers than those in the other groups. The two groups made up of women, HPP and EP, reported improvements in household food consumption as well as some aspects of women’s empowerment. In the HPP group, households consumed greater amounts of fish and vegetables, while EP households regularly consumed two or three meals every day. Increased income among both groups led to better household financial security and in women’s decision-making power regarding the additional funds. EP groups, in particular mentioned using the income for children’s education expenses. Further, EP beneficiaries greater support from their husbands and other men, while HPP noticed increased interest and participation of men and children in homestead gardening. Access to capital and cash for expansion was the largest obstacles for these groups, and both HPP and EP group discussed the need for financial assistance to invest and expand operations. An additional recommendation was to expand training opportunities and activities and the include training for men. Participants suggested that inclusion of male family members would also improve project sustainability. Most PP participants were male and all were established farmers who met household consumption needs. This group focused on improving value chain linkages and improving their farm business. The PP beneficiaries discussed changing their mindset regarding to understand modern and appropriate techniques. The female participants in the group noted their increased 83 role in production and marketing of fish and vegetables. While this group reported gaining skills and knowledge, they still felt a need for support and guidance from the program and were hesitant to use resources for adopting new technologies. As with the other groups, PP members recommended further training, as well as program support for adopting technologies and practices. Regarding sustainability, the PP group said that continuing the program for two or three more years would greater improve their ability to expand and diversify to a point where they would not require external assistance. SO3 – Disaster Risk Reduction Through SO3, Nobo Jibon sought to provide greater protection for children and their families through contingency planning and improved emergency response. Both baseline and endline surveys included questions related to behaviors during past disasters, natural disaster preparedness, and ability to resume livelihood activities in the wake of recent disasters. Table 37 presents the results. Notably, Table 37 shows that, among all surveyed households and across each district, a significantly greater number of households had a plan to protect members, livestock, or assets in the event of a disaster compared to baseline (40 percent increase). Nearly three-quarters of households report having disaster plans in Barguna and Patuakhali (74 percent and 77 percent, respectively). Growth in households reporting disaster plans was greatest in Barisal district, increasing from 26 percent at baseline to 46 percent at endline. Table 37: Household preparedness and impact of recent disaster, percentage by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Households with a plan to protect members, livestock, or assets in the event of a disaster All households 45.8 64.1 39.8 * 5,026 5,346 District         Barisal 26.3 45.5 72.9 * 1,649 2,019      Barguna 56.1 74.1 32.1 * 1,565 1,615      Patuakhali 54.8 76.6 39.8 * 1,812 1,712 Households with loss of life during last disaster All households 0.6 0.5 ‐15.8 5,026 5,160 District         Barisal 0.4 0.2 ‐43.5 1,649 1,860 84 Table 37: Household preparedness and impact of recent disaster, percentage by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline      Barguna 1.3 1 ‐23.4 1,565 1,607      Patuakhali 0.3 0.5 54.4 1,812 1,693 Minimal asset loss in last disaster All households 3.8 3.8 0.0 5,026 4,415 District         Barisal 4.9 7.2 46.5 * 1,649 1,370      Barguna 0.5 2.3 333.3 * 1,565 1,460      Patuakhali 5.6 2.2 ‐60.4 * 1,812 1,584 Able to resume livelihood activities within 2 weeks following a natural disaster   All households 73.8 80 8.4 * 5,026 5,160 District         Barisal 75.2 81.5 8.4 * 1,649 1,860      Barguna 72.5 84.5 16.5 * 1,565 1,607      Patuakhali 73.8 74.2 0.6 1,812 1,693 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Few households reported minimal asset loss in the last disaster (four percent), though every district experienced significant, albeit mixed, changes. The endline totals reflect significant increases in Barisal and Barguna and a significant decrease in Patuakhali. Households in Barisal were most likely to experience minimal loss (seven percent), while just two percent of households in the other districts reported this. No significant change was seen in the percentage of households with loss of life during the last disaster. More notably, four out of five households (80 percent) were able to resume livelihood activities within two weeks following a natural disaster, up from 74 percent at baseline. Both Barisal and Barguna districts reported significant gains in this area (eight percent and 17 percent, respectively). Nearly half of all households (48 percent) received warning within 12 hours of the last disaster up from 37 percent at baseline (Figure 13). Barguna district experienced the largest increase from 38 percent of households to 62 percent. A significant increase was also observed in Patuakhali, while 27 percent of Barisal households received warning, down from 30 percent. 85 Figure 13: Households who received warning within 12 hours of the last disaster Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) The proportion of households that received training in disaster preparedness increased markedly, overall, as shown in Figure 14. At endline, 22 percent of households had received training compared to five percent at baseline. Similarly large gains were seen in each district. The largest increase in trained households was in Barguna district, where 32 percent reported training (six percent at baseline). Likewise, Patuakhali improved from seven percent to 24 percent, and Barisal increased from one percent to 12 percent. Figure 14: Households who received disaster preparedness training, by district Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 10% 27% 36% 25% 6% 41% 45% 30% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Barisal* Barguna* Patuakhali* Total Baseline (n: 5,025) Endline (n: 5,160) 1% 6% 7% 5% 12% 32% 24% 22% 0% 5% 10% 15% 20% 25% 30% 35% Barisal* Barguna* Patuakhali* Total Baseline (n: 5,026) Endline (n: 5,346) 86 Disaster response also improved among all households, rising from 25 percent to 30 percent (Figure 15). This improvement is reflected in both Barguna and Patuakhali districts, which reported 41 percent and 45 percent of households, respectively, that sought shelter (27 percent and 36 percent, respectively, at baseline). Barisal, in contrast reported a small but significant decrease from 10 percent to six percent of households. Figure 15: Households that sought shelter within 12 hours of the last disaster Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Qualitative Information (SO3) Activities in this SO were split between youth volunteers and adults. After receiving DRR training, 25-member youth committees with equal numbers of boys and girls organized courtyard sessions and trainings for community members. Youth volunteers reported using their knowledge, without incentive, to actively engage the community in skill-building around all aspects of disaster management. This group offered two recommendations: first, stronger linkages to the union disaster management committee (UDMC) and, second, additional training and equipment, including a first aid kit. Separate groups were formed for men and women, who received a DRR orientation and participated in courtyard sessions held by the youth. Adults stated that the topics covered in the trainings were useful and every household has a written contingency plan. Participants were better prepared for a disaster and followed the information in their plans. Some families preferred to stay with friends or neighbors rather than at a cyclone shelter, either because not enough shelters were available or because they feared gender or socio-cultural discrimination. Adults recommended building more shelters and improving infrastructure, as well as continued public and private support and improved gender equity and women’s empowerment. Project Participation 10% 27% 36% 25% 6% 41% 45% 30% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Barisal* Barguna* Patuakhali* Total Baseline (n: 5,025) Endline (n: 5,160) 87 In order to assess the extent to which project interventions contributed to changes in outcome and impact indicators, comparisons of these higher level indicators are made across households that reported participating in different combinations of project interventions, focusing on participation in SO1 and SO2 interventions. The evaluation was not designed to provide statistically representative comparisons between participating and non-participating households or to be able to measure the extent to which observed differences between participants and non￾participants are attributable to project interventions. The following analysis takes advantage of the fact that information was collected from both participants and non-participants, to measure differences in outcomes at endline between households that directly received Nobo Jibon services and those that did not. Any differences found should not be interpreted as being indisputably attributable to program impact; however, (positive) differences should be considered along with other information in this report as providing supporting evidence that Nobo Jibon is achieving program objectives. Results from the endline QPE also provide information about the extent to which project interventions were targeted toward more food insecure households. Examination of participation in the project interventions by food security category (Table 38) shows that the most food￾insecure households participated marginally more in all types of project interventions (SO1, SO2 and SO3) than households in the higher food security categories. However, there is not much difference in participation in either SO1 or SO3 across the food security categories. This suggests that there is no strong targeting toward food-insecure households for these intervention areas, as we are not seeing any differences in participation in these program areas between highly food insecure households and low food insecure households. This result is consistent with the overall programming strategies for these two SOs; SO1 support is available to all pregnant women and mothers of young children regardless of their food security status, while interventions under SO3 are intended to benefit all households within a supported community. The results of participation by food security status under SO2 provide some evidence of targeting, as a higher proportion of households in the lowest food security category participated in SO2 activities than those in higher categories. Again, this is consistent with the project strategy, in which these interventions are generally targeted toward more food-insecure populations. However, the project also directed some types of support to (more food-secure) larger farmers, as a means to enhance marketing opportunities and demand for agricultural labor for all households within communities. 88 89 Table 38: Households participation in SO1, SO2 & SO3; by food security category    % HH participating Food Security Category SO1 SO2 SO3    1 Lowest 50.8 27.1 63.3    2 Middle 47.2 * 23.5 * 57.8 * 3 Highest 47.8 21.7 * 57.8 * Total sample 48.6 24.1 59.7     Note:  Stars (*) for program participation across food security categories indicate difference is statistically significant at the p<.10 level when compared to low food security at endline. The following tables provide information on overall and severe child malnutrition, HFIAS, CSI, HDDS, and MAHFP disaggregated by program participation. When disaggregated by program participation, program goal indicators show limited significance (Table 39). The table compares household that received no program assistance to households that received SO1 programming only, SO2 programming only, or a combination of SO1 and SO2 programming. From baseline to endline, all households showed improvement in overall stunting among children 6-59 months (20 percent decrease, Table 39). However, there were no statistically significant differences in overall stunting rates between the different program participation categories at endline. Table 39: Key program goal indicators, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline % of overall stunted (HAZ<‐2SD) children age 6‐59 months All households 43.9 35.3 ‐19.6 + 2,296 1,853 Program participation          Did not participate SO1 or SO2 30.5 ‐30.5 308       Participated SO1 only 35.5 ‐19.1 1,037       Participated SO2 only 36.0 ‐18.0 28       Participated SO1 & SO2 38.0 ‐13.4 480 Household Food Insecurity Access Scale (HFIAS)   All households 28.7 19.4 ‐32.4 + 5,009 5,346 Program participation          Did not participate SO1 or SO2 17.9 ‐37.6 2,377       Participated SO1 only 18.8 ‐34.7 1,677       Participated SO2 only 27.5 ‐4.3 * 351 90 Table 39: Key program goal indicators, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline       Participated SO1 & SO2 21.4 ‐25.5 * 941 Coping Strategy Index All households 13.5 8.4 ‐37.8 + 4,969 5,346 Program participation          Did not participate SO1 or SO2 7.4 ‐44.7 2,377       Participated SO1 only 8.1 ‐39.7 1,677       Participated SO2 only 12.7 ‐5.8 * 351       Participated SO1 & SO2 9.6 ‐28.9 * 941 Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not receive SO1 or SO2" at endline. Likewise, improvements were reported among all households in food insecurity and coping strategies (Table 39). Overall, households scored lower on the HFIAS (19.4) than at baseline (28.7), a 32 percent decrease. Similarly, across the sample a 38 percent drop was seen in the CSI from 13.5 to 8.4. There were no statistically significant differences between households that participated in SO1 and those households that did not participate in either SO1 or SO2 in the average value of the HFIAS or CSI indices. HFIAS at endline (Table 39) was worse for households that received either only SO2 (27.5) or both SO1 and SO2 (21.4) than for those households that did not participate in SO1 or SO2 (17.9). Likewise, for the CSI index at endline, households that received SO2 in any form (SO2 only: 12.7; SO1 and SO2: 9.6) had worse levels of the CSI index that non-participant households (no SO1 or SO2: 7.4). These results may be reflective of the targeting, discussed above, with respect to SO2 programming. Households receiving SO2 programming are likely to be worse off than household not receiving SO2 programming across a range of measures (see Table 3 and Table 38). Because this is true, the results between households that received SO2 and those that did not receive any programming are not exactly comparable. Table 40 shows SO2 impact indicators (HDDS and MAHFP) by program participation. Both of these measures improved across all sample households, with very little variation among program participation categories. As is the case with the program goal indicators, there was no statistically significant difference at endline between households that received SO1 (HDDS: 5.7, MAHFP: 10.4) and those that did not participate in SO1 or SO2 (HDDS: 5.7, MAHFP: 10.5) in these indicators. SO2 only households performed marginally worse (HDDS: 5.4, MAHFP: 9.9) and MAHFP than households that did not receive SO1 or SO2. Again, the relatively poorer 91 performance of SO2 only households is likely more a reflection of program targeting than of program (in)effectiveness. Table 40: SO2 impact indicators, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Household Dietary Diversity Score (HDDS) All households 4.7 5.7 20.8 + 5,026 5,346 Program participation          Did not receive SO1 or SO2 5.7 20.9 2,377       Received SO1 only 5.7 21.9 1,677       Received SO2 only 5.4 15.5 * 351       Received SO1 & SO2 5.7 20.6 941 Months of Adequate Household Food Provisions (MAHFP) All households 9.4 10.4 10.2 + 5,026 5,346 Program participation          Did not receive SO1 or SO2 10.5 11.4 2,377       Received SO1 only 10.4 10.5 1,677       Received SO2 only 9.9 5.1 * 351       Received SO1 & SO2 10.2 8.5 * 941 Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not receive SO1 or SO2" at endline. Data on three childhood feeding practices disaggregated by participation in SO1 are shown in Table 41. Among all surveyed households, significant gains are seen in U2 children who receive a minimally acceptable diet (290 percent increase). However, there is no significant difference at endline in the proportion of households with minimally acceptable diet between those households that received SO1 (23 percent) and those that did not receive SO1 (20 percent). There is no significant change reported in the proportion of U5 children exclusively breastfed for all households sampled from baseline to endline ( 92 Table 41). Likewise, there is no difference in rates of exclusive breastfeeding between households that received SO1 (44 percent) and those that did not receive SO1 (46 percent). Table 41: Childhood feeding practices, by program participation in SO1   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline Infants/toddlers 6‐23  months who receive a minimally acceptable diet (apart from breast milk) All households 5.8 22.6 289.7 + 793 688 Program participation          Participated SO1 23.3    551       Did not participate SO1 19.5    137 Children under 6 month exclusively breastfed All households 38.4 44.9 16.9 282 320 Program participation          Participated SO1 44.4    227       Did not participate SO1 46.3    93 Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not receive SO1 or SO2" at endline. When comparing across program participation for SO2 activities for improved agricultural techniques, the data in Table 42 shows significant differences between participants and non￾participants across both measures. The percent of households adopting improved marketing practices were higher for SO2 participants at 2.5 percent compared to non-participants at 1.6 percent. For the percent of household adopting three or more improved agricultural practices, participant households reported a 9.5 percent adoption rate compared to a 5.9 percent adoption rate for non-participants. Table 42: Use of improved agricultural techniques, by program participation in SO2   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % HH adopting improved marketing practices All households 0.4 1.9 324.7 + 1,184   1,825   Program participation          Participated SO2 2.5    * 509         Did not participate SO2 1.6    1,317   % HH adopting 3 or more improved agricultural practices 93 All households 4.9 6.9 42.2 + 2,025   3,071   Program participation          Participated SO2 9.5    * 811         Did not participate SO2 5.9    2,260   Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not participate SO2" at endline. Adoption of agricultural practices shown in Table 43 reveal that the adoption of animal manure, crop rotation, and fertilizer was higher for project participants than non-participant households. Participants, at endline were 22.4 percent more likely to use animal manure, 50.0 percent more likely to use crop rotation, and 3.9 percent more likely to use fertilizers. The data also show that a smaller proportion of participants have access to improved irrigation techniques compared with non-participants. at nearly half the rate (48 percent) of non-participants. It should be emphasized that the overall level of improved irrigation use is extremely low for all households in the sample (1.5 percent and 2.9 percent for participants and non-participants, respectively). Table 43: Use of improved agricultural techniques, by technique and program participation (endline)   Indicator Endline Percent difference (Participant‐        Non‐participant) Number of observations % HH adopting improved agricultural techniques (endline)    Animal manure       Participated SO2 33.9 22.4 * 811       Did not participate SO2 27.7 2,260 Compost       Participated SO2 34.5 8.8 811       Did not participate SO2 31.7 2,260 Crop rotation       Participated SO2 13.8 50.0 * 811       Did not participate SO2 9.2 2,260 Fertilizer       Participated SO2 83.6 3.9 * 811       Did not participate SO2 80.5 2,260 Biological pest control       Participated SO2 10.0 ‐2.9 811       Did not participate SO2 10.3 2,260 Mechanical pest control       Participated SO2 3.2 ‐11.1 811 94 Table 43: Use of improved agricultural techniques, by technique and program participation (endline)   Indicator Endline Percent difference (Participant‐        Non‐participant) Number of observations       Did not participate SO2 3.6 2,260 Chemical pest control       Participated SO2 69.8 4.0 811       Did not participate SO2 67.1 2,260 Integrated pest management       Participated SO2 11.6 18.4 811       Did not participate SO2 9.8 2,260 Improved irrigation       Participated SO2 1.5 ‐48.3 * 811       Did not participate SO2 2.9 2,260 Note:   Stars (*) for program participation indicate difference is statistically significant compared to "did not participate SO2" at endline. Two indicators for disaster preparedness and response are shown in Table 44. Overall, the proportion of households with a disaster plan increased from 46 percent to 64 percent, a 40 percent increase. Households that participated in SO3 are much more likely to have a disaster plan (70 percent) compared to those that did not participate in SO3 (56 percent). Four-fifths of households (80 percent) were able to resume livelihoods activities within two weeks of a natural disaster, an eight percent increase. Those households that did not participate in SO3 improved more than those that did (82 percent versus 79 percent, respectively). However, the magnitude of the difference is small. Table 44: Household preparedness and impact of recent disaster, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline %HH with a plan to protect members, livestock, or assets in the event of a disaster All households 45.8 64.1 39.8 +        5,026          5,346   Program participation          Participated SO3 69.6    *        3,203       Did not participate SO3 55.6           2,143 % HH able to resume livelihood activities within 2 weeks following a natural disaster   All households 73.8 80.0 8.4 +        5,026          5,160   Program participation    95 Table 44: Household preparedness and impact of recent disaster, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       Participated SO3 78.9    *        3,093         Did not participate SO3 81.8           2,067   Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not participate SO3" at endline. Vulnerable Groups Women’s Decision Making and Empowerment Nobo Jibon was designed to address two main causes of food insecurity in the program area: i) erratic and low-paying income earning opportunities, especially for asset-poor households, and ii) social exclusion and low status of women and children. Nobo Jibon thus aimed to strengthen the enabling environment for income generation and improved household economies, and to promote women’s engagement in household decisions. To assess progress in these indicators, both surveys asked women who earned cash by working on a regular basis outside the home about the source of their income. Additionally, female adult respondents who worked on a regular basis were asked to rate their level of participation in five common household decisions. Women were considered to have a voice in a decision if they could make it alone or jointly with their husband. Table 45 reports on the types of income-earning activities women engage in and on the types of household decisions they report making, disaggregated by food security category. Data were collected only for the households with women who report earning income. Raising poultry is by far the most common enterprise, with four-fifths (80 percent) of all households with income￾earning women engaged in poultry activities. Participation in all other activity types was very limited – under 10 percent for all activities listed on the survey. The one exception to this is the lowest food security tercile, where about 15 percent worked for daily wages. In terms of decision-making, at baseline, a solid majority of income-earning women – both sample-wide and within each food security category – reported being able to make decisions either alone or with her husband, for all decision types listed on the survey. This tendency was even stronger at endline, with three-quarters or more of income-earning women reporting decision-making authority for all decision types. The most marked change in this respect was for decisions on children’s health expenditures, which increased from 77 percent of the overall 96 sample to 87 percent, with larger increases for the low and medium food security groups. The evaluation data and methodology do not allow us to definitively attribute these changes to program efforts. The women’s economic empowerment score is the sum of scores for the five individual decisions (e.g. family visit decision making, children’s heal expenditure decision making, etc.). If the response for one of the five individual decision making questions indicated that a woman made a decision alone, or jointly with her husband, the score value for that particular component is one. If the response indicated that the decision was made by her husband, somebody else, or her husband and somebody else, the score value is zero. The summation of these 5 scores is the women’s empowerment score, with a maximum of 5 and a minimum of 0. The women’s economic empowerment score increased slightly, but significantly, from baseline to endline, from 3.7 to 4.2, out of a maximum score of 5.0. Table 45: Women's income earning activities and decision making, by food security category   Indicator Low Middle High Total N 1778 1779 1779 5336 Percent of all HH with a woman who earns income 38.8 30.9 29.9 33.2 % women's participation in income‐earning activities (endline) N 690 549 532 1771 Poultry 74.7 85.9 80.6 80.0 Daily wage earner 14.5 5.1 2.1 7.8 Agri/Farmer 6.7 7.4 6.0 6.7 Handicrafts/Handloom 8.2 4.3 4.8 6.0 Other 6.9 3.9 3.8 5.0 Services 1.0 2.4 7.7 3.5 Work in other household 6.9 1.3 0.0 3.1 Business 1.4 1.7 1.5 1.5 Private tutor 1.0 1.0 1.9 1.3 % women making household decisions (endline) N 690 549 532 1771 Family visits 77.3 73.5 79.7 76.8 Expenditures on children's health 88.3 84.2 86.6 86.5 How to spend women's income 80.9 78.9 88.2 82.5 Major household purchases 78.1 71.7 78.7 73.7 Purchases of daily household needs 80.5 76.4 82.9 76.3 Women's economic empowerment (mean, endline) Women's economic empowerment score 4.2 4.0 4.3 4.2 97 Women were further analyzed by their level of empowerment (Table 46). Women who scored 5.0 across the sum score for individual decision-making were considered more empowered, whereas women with scores less than 5.0 were considered less empowered. That is to say, those women that fall into the more empowered category, with a score of 5.0, are fully empowered with respect to decision making in all 5 categories of household welfare measured. Conversely, those that are less empowered indicated that in at least one of the categories of household decision making measured, that particular female respondent was not empowered to make a decision in at least one of the respective categories. It was found that women were significantly more empowered at endline (67.5 percent) than women at the baseline (56.3 percent). Although the trend of empowerment has gone up, it is not clear if this was related to participation in program activities especially when compared across program participation in SO1 and SO2 interventions (Table 46). Women eligible to participate in SO1 courtyard sessions (PLW and mothers with children under two years) were significantly less so empowered at endline (65 percent) than those women who did not participate in these interventions (70 percent). From focus group discussions, issues related to women empowerment were not emphasized by respondents as a key topics covered in these sessions; rather there was more focus on pregnancy and child care practices. Women’s empowerment issues under SO1 were more so highlighted with the program’s adolescent groups; topics included gender inequality, early marriage, mobility, and violence against women. Since adolescents were not included in the evaluation design, data was not available to gauge their level of empowerment stemming from program participation in SO1 interventions. For SO2, despite having more empowerment and income generation-focused interventions, participants showed no significant difference when compared to non-participants. Table 46: Women’s decision making and empowerment, by participation   Indicators Baseline Endline Percent difference p‐ value Women's decision making score (mean, baseline, endline) N 1519 1774 Women's decision making score a/ 3.7 4.2 11.3 + Women's empowerment (endline) % women more empowered 56.3 67.5 19.9 + % women more empowered, by participation in SO1 interventions b/      Participant ‐ 64.9 ‐7.5 *      Non‐participant ‐ 70.2 % women more empowered, by participation in SO2 interventions b/ 98   Indicators Baseline Endline Percent difference p‐ value      Participant  ‐  67.9 0.7      Non‐participant  ‐  67.4 Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not receive SO1 or SO2" at endline.   a/ Percent difference and p‐value are based on the mean difference between the endline and baseline measurements. b/ Percent difference and p‐value are based on the difference between non‐ participants and participants in program specific interventions Child Rights and Protection Village Development Committees (VDC) are a central aspect of Nobo Jibon. They aid in consciousness-raising about legal rights, campaign and network to protect human rights, and mitigate domestic conflicts. One of the aspects through which the effect of VDC’s role was measured is household awareness and beliefs about child protection issues. Figure 16 displays the results. Figure 16 shows increases in parents’ understanding of child rights. More than half of households reported an awareness of children’s rights to education and health services. Significant increases were also seen in acknowledgement of rights to live with parents and to give an opinion, though less the one-quarter of households reported this. In addition, the percentage of parents who did not know any rights of children decreased from more than one￾third to less than one-quarter. Figure 16: Reported rights of children acknowledged by parents 99 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Figure 17 presents interesting findings with respect to the question of whether parents believe that hitting children when they have done something bad is wrong. At endline, fewer parents agree that this is wrong (73 percent compared to 83 percent at baseline), and more parents disagree with the idea (25 percent versus 15 percent). Both differences were significant, suggesting that more parents believe it is OK to hit a child when they have done something bad. Figure 17: Percentage of responses to the question: "Is it wrong to hit children whenever they do something bad?" Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Figure 18 shows several conditions and whether parents believe children should be protected from these. Nearly three-quarters of parents understood that children have the right to be 0% 10% 20% 30% 40% 50% 60% 70% Other Protection from abusive child… Birth registration Protection from… Give opinion* Recreation Don't know* Live with parents* Health services* Education* Endline (n: 5,346) Baseline (n: 5,026) 82.8% 15.0% 2.2% 73.2% 25.2% 1.6% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% Agree* Disagree* Don't know* Baseline (n: 5,009) Endline (n: 5,345) 100 protected from physical natural threats at endline, compared to less than half of parents at baseline. Acknowledgement of protection from physical abuse also increase slightly, though remained at less than one-quarter of parents. Less than 20 percent of parents reported that they did not know any condition from which children need protection, down from almost 40 percent at baseline. Other areas where little or no change was reported include early marriage, abusive child labor, trafficking, social stigma, and sexual abuse. Figure 18: Percentage of responses to the question: "What children should be protected from?" Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 0% 10% 20% 30% 40% 50% 60% 70% 80% Other Sexual abuse Social stigma Trafficking* Abusive child labor* Early marriage Don't Know* Physical abuse* Physical/Natural threats* Endline (n: 5,346) Baseline (n: 5,026) 101 4. Conclusions The purpose of the final QPE is to measure changes in project impact and outcome indicators over the life of the Nobo Jibon project, in order to assess the extent to which project objectives have been achieved, measure the overall impacts on populations in the project areas, assess the assumed causal pathways linking project activities to outcomes and impacts, and determine how interventions contributed to achieving project goals. Comparison of baseline with endline values demonstrates that the Nobo Jibon project surpassed targets for all SO1 and SO2 impact indicators measuring household nutrition and food security status. In particular, the endline values for all anthropometric indicators, HFIAS, CSI, HDDS, MAHFP exceeded the target values for these indicators. The results for the SO3 impact indicators are favorable, as well; the percent of households with disaster preparedness plans, households trained in disaster preparedness, the percent of households that sought shelter in a timely manner, and those that received warning within with adequate lead time all increased from baseline to endline. However, none of these SO3 impact indicators met their targets. In addition, substantial improvements in project outcome indicators were recorded, with improvements in indicators measuring knowledge and adoption of recommended practices increased, although generally the targeted values were not achieved for many of the outcome indicators. The percent of households reporting adoption of recommended IYCF practices and other child care practices, child caregiver practices, diets and treatments of PLWs all increased from baseline to endline survey rounds. These changes in practices are consistent with the dramatic improvements in the SO1 and SO2 outcome indicators and suggest that the assumed causal relationships between outcome and impact indicators built into the design of the Nobo Jibon project are valid. The results also suggest that the targets set for the outcome indicators were perhaps overly ambitious, since the impact level goals were achieved even though the target values of the outcome indicators generally were not met. Looking at differences in outcomes in relation to participation in project activities, the extent to which changes in outcomes can be attributed directly to project interventions is generally not clear cut, and changes vary across the three SOs. In SO1, while there is generally a large increase in adoption of recommended practices from baseline to endline, there is little difference in adoption between households that participated in SO1 activities and those that did not participate at endline. These results point to a general adoption of improved behaviors, that may be a result of efforts of interventions of government programs or those of other organizations, or that the messages promoted by Nobo Jibon were effectively transmitted to individuals in the project area who did not participate directly in project interventions. In the case of SO2, the relationship between participation and adoption of practices is a bit stronger. A higher proportion of SO2 participants adopted improved agricultural production and 102 marketing practices. However, the overall proportion of sampled households that adopted these improved practices was quite small, even in the endline round. Results from the endline QPE also provide information about the extent to which project interventions were targeted toward more food insecure households. Examination of participation in the project interventions by food security category shows that the most food-insecure households participated more in all types of project interventions (SO1, SO2 and SO3) than households in the higher food security categories. The relative proportions of project participation across the food security categories, however, are not very pronounced for either SO1 or SO2, suggesting that there is no strong targeting toward food-insecure households for these intervention areas. This result is consistent with the overall programming strategies for these two SOs; SO1 support is available to all pregnant women and mothers of young children regardless of their food security status, while interventions under SO3 are intended to benefit all households within a supported community. The results of participation by food security status under SO2 provide some evidence of targeting, as a higher proportion of households in the lowest food security category participated in SO2 activities than those in higher categories. Again, this is consistent with the project strategy, in which these interventions are generally targeted toward more food-insecure populations. However, the project also directed some types of support to (more food-secure) larger farmers, as a means to enhance marketing opportunities and demand for agricultural labor for all households within communities. One important thrust of the programming strategy of Nobo Jibon has been to reduce the exclusion of women and other vulnerable groups (especially children) from economic and social opportunities and to enhance the economic empowerment of women. According to information collected from women who had access to income, their economic empowerment, as measured by decision-making authority over income and economic activities, has increased from baseline to endline. However, it is important to note that this increase has occurred to a very similar extent in households that have not participated in project activities as compared with project participant households. One important finding from the qualitative research is that the project interventions with youth seem to have a strong and long-term impact on empowering girls and women. Important implications from this finding are that i) programming strategies directed toward youth may have strong impacts on enhancing empowerment of women, and ii) indicators of empowerment should be measured on youth. 103 5. Recommendations Given the strong ongoing investment in health and nutrition programming by government and other non-government sources, taken together with strong gains in health and nutrition observed in the program area over the course of the current round of MYAPS in Bangladesh, it is now appropriate to review the level of support provided for mother and child health, and nutrition (MCHN) programming by FFP resources, to minimize overlap with complementary offerings by other organizations and the government of Bangladesh (GOB) and consolidate these services to eliminate any possible redundancies. In particular, the results from this QPE of Nobo Jibon suggest that future FFP programing resources and efforts in Bangladesh could be usefully diverted towards programming to support enhancing livelihoods, where the impacts on project participants have been great, but the reach of program interventions has been limited in terms of the number of participants. Another recommendation refers to future project monitoring and evaluation design. One very notable limitation of the Nobo Jibon final project QPE was that the scope of work was essentially restricted to a quantitative household survey, following a performance monitoring design based on a population-based sample design. With a clear appreciation of this limitation in the scope of the QPE, Save the Children included a small qualitative component in the scope of work for the final evaluation. However, the scale of this exercise was very restricted, and in particular was limited to only collecting information from project beneficiaries. There was no scope within the terms of reference for interviews with project staff, other implementing organizations or other stakeholders to get any qualitative detail about project implementation, or assessments about the strengths and weaknesses of project design and implementation with respect to achieving impacts with beneficiaries. The combination of the performance monitoring design, with the separation of the quantitative final QPE from the overall qualitative evaluation of FFP projects in Bangladesh made interpretation of the quantitative results very difficult. In the future, project M&E plans should include an integrated final project evaluation design that includes both qualitative and quantitative components. Furthermore, FFP should ensure that plans and resources for more carefully designed impact evaluations are included within project or country￾level plans, so that studies to measure the contribution of project activities to measured outcomes and impacts can be undertaken. 106 Annex 2: Mean Values and Confidence Intervals for Indicator Performance Tracking Table (IPTT) Indicators Indicator Indicator Type Baseline3 Endline 95% C.I. (Endline) LOA Target Goal: Reduced food insecurity and vulnerability for 191,000 households (direct beneficiaries) in ten Upazilas of Barisal Division in southern Bangladesh over five years Percentage of stunted (HAZ<-2) children aged 6-59 months 1 <- 2SD Impact 43.90% 35.30% 33.3 - 38.2% 39.50% <- 3SD 12.90% 10.00% 8.3-11.7% 11.00% Average HH Food Insecurity Access Scale score Impact 28.70% 19.40% 18.3 - 20.5% 25.80% Average HH coping strategy index Impact 13.50% 8.40% 7.8 - 8.9% 12.20% SO1 MCHN: Improved health and nutritional status of children U5 and PLW Percentage of underweight (WAZ<-2) children aged 0-59 months 1 <- 2SD Impact 39.40% 27.30% 25.4 - 29.9% 35.50% 107 <- 3SD 9.90% 5.20% 4.0 - 5.8% 8.40% Percentage of underweight (WAZ<-2) children aged 0-23 months 1 <- 2SD Outcome 31.90% 19.50% 15.9 - 21.7% 28.80% <- 3SD 7.60% 4.50% 2.9 - 5.7% 6.90% Percentage of wasted (WHZ<-2) children aged 6-59 months 1 <- 2SD Impact 15.90% 11.00% 9.2 - 12.4% 14.30% <- 3SD 2.00% 1.40% 1.5-2.6% 1.70% Percentage of wasted (WHZ<-2) children aged 6-23 months 1 <- 2SD Outcome 15.10% 13.80% 10.0-16.2% 13.60% <- 3SD 3.00% 3.00% 1.5-4.2% 2.30% IR 1.1.: PLW and care-givers of children U5 practice improved MCHN and environmental health behaviors % of infants 0-5 months of age who are fed exclusively with breast milk2 Outcome 38.40% 44.90% 39.0 - 50.9% 65.00% 108 % of children 6-23 months of age who receive a minimum acceptable diet (apart from breastmilk)2 Outcome 5.80% 22.50% 19.1 - 26.1% 25.00% % of caregivers demonstrating proper personal hygiene behaviors Outcome 30.90% 38.10% 34.9 - 41.0% 50.00% % of beneficiary caregivers demonstrating food hygiene behaviors Outcome 20.20% 26.60% 23.9 - 29.2% 50.00% % of PLW who consume food rich in iron Outcome 31.50% 90.50% 87.9 - 93.2% 60.00% % of PLW who consume food rich in Vitamin A Outcome 22.30% 59.60% 55.2 - 64.0% 60.00% % of PLW who consume food rich in Calcium Outcome 12.20% 12.40% 8.8 - 16.0% 40.00% % of PLW taking iron or iron folate supplements in the last 7 days Outcome 2.10% 11.80% 8.5 - 15.1% 50.00% IR 1.2.: Households have improved access to integrated health, family planning and nutrition services % of children 12-23 months who received Vitamin-A supplementation in the past 6 months Outcome 42.30% 45.40% 37.4 - 47.3% 85.00% 109 % of mothers of children aged 6-23 months who received high-dose Vitamin A supplement within 8 weeks postpartum (6 weeks if not exclusively breastfeeding) in last pregnancy Outcome 21.00% 38.20% 33.9 – 42.4% 50.00% % of mothers attended ANC session at least 4 times during last pregnancy Outcome 11.80% 32.90% 29.3% - 36.5% 50.00% % of beneficiary children 12-24 months receiving antehelminth (deworming) medication in previous 6 months Outcome 18.80% 32.80% 28.0 – 37.6% 30.00% IR 1.3. : Equity increased within households and communities % of beneficiary women whose husband attends ANC/PNC with her5 Outcome 48.60% 40.40% 36.4 - 44.4% 50.00% SO2 Market-based Production and Income Generation: Poor and extremely poor households have increased production and income Average HH dietary diversity score (HDDS) Impact 4.7 5.7 5.6 - 5.7 5.5 Average HH dietary diversity score (HDDS) Outcome 4.7 5.7 5.6 - 5.7 5.5 110 Average number of months of adequate household food provisioning (MAHFP) Impact 9.4 10.4 10.3 -10.5 11 % of HHs reporting increase in production of one or more products Outcome 38.80% 43.60% 41.5% - 45.7% 50.00% Average annual income from sale of agricultural products Outcome 5,823 10,628 9,832 - 13,632 12,950 IR 2.1.: Poor households apply improved knowledge and skills for production and marketing %of beneficiaries (farmers) using 3 or more sustainable/improved production practices. Outcome 4.80% 9.50% 7.5 – 12.0% 20.00% % of targeted HHs adopting improved marketing practices Outcome 0.00% 2.50% 1.0% - 4.6% 70.00% SO3 DRR: Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters % of HHs with a feasible plan to protect human life and productive assets during disaster Impact 45.90% 64.10% 61.9% - 66.0% 75.00% 111 %of HHs able to resume livelihood activities within 2 weeks following a natural disaster. Impact 73.80% 80.00% 78.2% - 81.9% 90.00% IR 3.1.: Communities manage functional emergency preparedness and response plans % of targeted HH members trained on disaster preparedness Output 4.60% 22.00% 19.8 - 24.3% 50.00% IR 3.4.: Communities receive and respond to early warning for floods and cyclones % of HHs that sought shelter in a timely manner during last disaster. Outcome 24.80% 29.60% 27.2% - 32.0% 50.00% % of HHs that received location specific cyclone warning signal with adequate lead time Output 0.00% 47.90% 45.7% - 50.0% 75.00% 112 Annex 3: Procedures for Computing Household Economic and Food Security Status Indicators 1. Asset Index This index is computed by multiplying the number of each type of household asset by the index value for that particular asset type. Index values of household assets used for construction of the asset index are presented in Table A 1. A higher value of the asset index indicates that households have been able to accumulate assets over time. Households are able to accumulate assets if income is greater than the necessary expenditures to meet household subsistence requirements. Assets also provide households with a cushion to adjust to shortfalls in incomes, or sudden increases in necessary expenditures. Thus, households with a higher asset index are less vulnerable than households with lower asset index values. Table A 1: Estimated average values (in USD) used in calculating household asset index Asset Index value Almirah 50 Table/chair/bench 10 Watch/clock 30 Cot/bed 20 Working radio 30 Working TV 100 Bicycle 100 Motorcycle 800 Phone 50 Rickshaw/van 300 2. Household Dietary Diversity Score (HDDS) This indicator is computed by summing the number of different food categories reported eaten by the household in day prior to the interview. This indicator was measured as recommended by FANTA, using the following 12 food groups: cereals, tubers, legumes, dairy, meat, fish, oils, sugar, fruits, eggs, vegetables, and others. The HDDS provides a measure of a particular household’s food access. A higher HDDS represents a more diverse diet, which is empirically highly correlated with a household’s income level and access to food.31 3. Months of Adequate Household Food Provisioning (MAHFP) 31 Swindale, Anne, and Paula Bilinsky. Household Dietary Diversity Score (HDDS) for Measurement of Household Food Access: Indicator Guide (v.2). Washington, D.C.: Food and Nutrition Technical Assistance Project, Academy for Educational Development, 2006. 113 This indicator reflects a household’s ability to obtain food from their own production, stocks, purchases, gathering, or through food transfers from relatives, members of the community, the government or donors. As a household manages its resources over the course of a year, the ability to meet its food needs may vary due to any number of factors such as inadequate crop production by the household due to poor soils or lack of labor, loss or decrease in income sources such as employment, social obligations or natural disaster. Measuring the MAHFP has the advantage of capturing the combined effects of a range of interventions and strategies, such as improved agricultural production, storage and interventions that increase the household’s purchasing power. 32 4. Household Food Insecurity Access Scale (HFIAS) This indicator has been developed by FANTA, and is based on household access to food and responses to shortages in access to food over a 30-day recall period. This indicator is based on the household’s: i) perceptions of uncertainty over food access in the past 30 days; ii) perceptions of insufficiency in quantity and quality of food over the past 30 days; iii) reported reductions in food intake; and iv) reported consequences of reductions in food intake. A higher value of this index indicates a higher degree of food insecurity. In tabulating the HFIAS score, a HFIAS score variable is calculated for each household by summing the codes for each frequency-of-occurrence question. The maximum score for a household is 27 (the household response to all nine frequency-of-occurrence questions was “often”, coded with response code of 3); the minimum score is 0 (the household responded “no” to all occurrence questions, frequency-of-occurrence questions were skipped by the interviewer, and subsequently coded as 0 by the data analyst.) The higher the score, the more food insecurity (access) the household experienced. The lower the score, the less food insecurity (access) a household experienced. 5. Coping Strategy Index (CSI) The coping strategy index is computed on the basis of a series of questions asked to respondents about how frequently they utilize a list of 12 possible strategies.33 The twelve strategies are the following: 1) Limit portion size at meal times 2) Reduce number of meals eaten per day? 3) Borrow food or rely on help from friends or relatives? 4) Rely on less expensive or less preferred foods? 32 Bilinsky, Paula, Anne Swindale. 2007. Months of Adequate Household Food Provisioning (MAHFP) for Measurement of Household Food Access: Indicator Guide. FANTA. June 2007. 33 Maxwell, Daniel, Richard Caldwell and Mark Langworthy. “ Measuring food insecurity: Can an indicator based on localized coping behaviors be used to compare across contexts?” Food Policy, Volume 33, Issue 6, December 2008 114 5) Purchase/borrow food on credit? 6) Gather unusual types or amounts of wild food / hunt? 7) Have household members eat at relatives or neighbors? 8) Reduce adult consumption so children can eat? 9) Rely on casual labor for food? 10) Abnormal migration for work 11) Skip entire day without eating 12) Consume seed stock to be saved for next season The frequency of adoption of each category is coded according to the following categories: 0 = never 1 = seldom 2 = sometimes 3 = often 4 = daily The coded frequency response for each strategy is then weighted by the severity weight of each strategy. Average severity weights across several coping strategies conducted in countries around the world34 are then applied to each coping strategy, using the following formula: CSI = Σ(frequency categoryi * severity weighti) i=1 to 12 The severity weights are as follows: Strategy Severity weight Limit portion size at meal times 2.3 Reduce number of meals eaten per day? 2.7 Borrow food or rely on help from friends or relatives? 2.5 Rely on less expensive or less preferred foods? 1.8 Purchase/borrow food on credit? 2.9 Gather unusual types or amounts of wild food / hunt? 2.9 Have household members eat at relatives or neighbors? 3.3 Reduce adult consumption so children can eat? 2.6 Rely on casual labor for food? 3.4 Abnormal migration for work 3.4 Skip entire day without eating 4.6 Consume seed stock to be saved for next season 3.6 115 6. Personal hygiene behavior Personal hygiene practices are based on the following appropriate hand washing behaviors Appropriate times to wash hands: 1. Before food preparation 2. Before eating 3. Before feeding children 4. After defecation 5. After cleaning babies bottoms Appropriate washing practices 6. Use water 7. Use soap or ash 8. Wash both hands 9. Rubs hands at least 3 times 10. Dries hands by air or with clean cloth “Proper personal hygiene behavior” is defined as following at least 8 out of these 10 practices (80%). Note that this is consistent with the definition used in the Jibon o Jibika baseline and end￾line surveys. 7. Food hygiene behaviors “Proper food hygiene behaviors” is defined as applying all three of the following practices: washing hands before food preparation, and washing hands before eating, washing hands before feeding children . 8. Water hygiene behaviors “Proper water hygiene behaviors ” is defined as all applying all three of the following three practices: water stored at home, drinking water stored in separate containers, and water is kept covered. 9. Environmental hygiene behaviors “Proper environmental hygiene behaviors” is defined as applying at least five of the six following practices: Use hygienic latrine (ring slab/offset latrine with water seal, covered open pit latrine, or septic latrine) Latrine is functioning Latrine shows signs of use Latrine (pan and slab) is clean Area surrounding latrine is clean Infants’ feces disposed of in latrines 116 10. Minimally acceptable diet A ‘minimum acceptable diet apart from breastmilk’ is calculated as follows: Breastfed children 6–23 months of age who had at least the minimum dietary diversity and the minimum meal frequency during the previous day and Non-breastfed children 6–23 months of age who received at least one milk feeding and had at least the minimum dietary diversity not including milk feeds and the minimum meal frequency during the previous day. This calculation differs slightly from that described in the World Health Organization’s guidelines for assessing and measuring infant and young child feeding practices (2008), which states non-breastfed children 6–23 months of age should receive at least two milk feedings and have at least the minimum dietary diversity not including milk feeds and the minimum meal frequency during the previous day. Minimum dietary diversity is defined as receiving four or more of the following foods: Rice, bread, porridge, other foods made from grain Tubers: white potatoes, white yams, other foods from roots Foods from beans, nuts, lentils Milk or milk products Liver, kidney, heart, fish, dried fish, seafood, any meat (chicken, beef, goat, duck, etc.) Eggs Pumpkin, carrots, orange sweet potatoes, dark green leafy vegetables, Ripe mangoes, ripe papayas, ripe jackfruits Any other fruits or vegetables 11. Economic empowerment index The scores for economic empowerment are calculated by taking the mean sum of scores for individual decisions. If the response indicated that a woman made a decision alone, or jointly with her husband, the score value is one. If the response indicated that the decision was made by her husband, somebody else, or her husband and somebody else, the score value is zero. The maximum score is five. 117 Annex 4: Results of Factor Analysis on Food Security Variables SPSS Factor Analysis Output: Factor Analysis The communalities table below shows how much of the variance in the individual elements that are accounted for in the factors extracted. For instance, a high proportion of the variance of HFIAS_and CSI indices (88.8 and 83.3 percent, respectively) are accounted for in the factors extracted in the analysis. Overall, all of the variables are well represented in the extracted factors, which ultimately will be used as the food security index. Communalities Initial Extraction HHsize 1.000 .720 assetindex_pc 1.000 .371 exp_month_pc 1.000 .698 food_share 1.000 .664 HDDS 1.000 .396 MAHFP 1.000 .617 HFIAS_index 1.000 .888 csi_index 1.000 .833 Extraction Method: Principal Component Analysis. Considering together the table and figure below that provide the total variance explained by the extracted factors and the scree plot that plots the eigenvalues of the extracted factors, one can see that the first factor accounts for a plurality of the cumulative variance and that after the first factor the additional variance and corresponding eigenvalues diminish rapidly. Using this information, only the first factor has been used for the food security index. 118 Total Variance Explained Component Initial Eigenvalues Extraction Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % 1 2.734 34.179 34.179 2.734 34.179 34.179 2 1.354 16.925 51.105 1.354 16.925 51.105 3 1.099 13.738 64.843 1.099 13.738 64.843 4 .928 11.602 76.445 5 .724 9.054 85.499 6 .582 7.270 92.769 7 .477 5.958 98.728 8 .102 1.272 100.000 Extraction Method: Principal Component Analysis. The component matrix below shows the loadings, or correlations, associated for each of the factors extracted. For the first factor, which is the only one used as part of the food security 119 index, the HFIAS index and CSI index have extremely strong positive correlations (.91 and .88, respectively) with the index and explain a good portion of the variation of the index. MAHFP and HDDS also contribute strongly to the index with correlations of -.77 and -.56, respectively (the correlations are negative because HDDS and MAHFP are scaled such that food security increases as these indicators increase, which is opposite scaling of the HFIAS and CSI index that decrease as food security improves. Component Matrixa Component 1 2 3 HHsize -.186 -.184 .807 assetindex_pc -.204 .287 -.497 exp_month_pc -.251 .797 -.010 food_share .302 -.695 -.298 HDDS -.560 .099 .270 MAHFP -.765 -.163 -.072 HFIAS_index .911 .206 .124 csi_index .880 .201 .136 Extraction Method: Principal Component Analysis. a. 3 components extracted. 120 Annex 5: Nobo Jibon Baseline Survey Household Questionnaire Nobo Jibon Baseline Survey Questionnaire (Quantitative Survey of Households) Questionnaire for Randomly Selected Households TANGO International and Save the Children-USA 121 INTRODUCTION AND CONSENT. Guidance for introducing yourself and the purpose of the interview:  Assalam walaikum/ Namashkar! My name is _____________ and I am currently working for/with the Save the Children Nobo Jibon Program on the Baseline Survey.  Your household has been selected by chance in this village for this interview. The purpose of this interview is to obtain information about the Livelihood, Maternal Child Health and Nutrition, Hygienic practices, disaster preparedness and responses and child protection. It will help us to understand the current status of the HH’s livelihood strategies in terms of scio-economic, health other related aspects.  The survey is voluntary and you/your family can choose not to take part. The information that you/your family give will be confidential. The information will be used to prepare reports, but will not include any specific names. There will be no way to identify that you gave this information.  Could you please spare some time (around 90 minutes) for the interview? NB to enumerator: DO NOT suggest in any way that household entitlements could depend on the outcome of the interview, as this will prejudice the answers. At this time, do you want to ask me anything about the survey? May I begin the interview now? RESPONDENT AGREES TO BE INTERVIEWED ........................................................................ 1  RESPONDENT DOES NOT AGREE TO BE INTERVIEWED ..................... 2 END Signature of interviewer: Date:_______________________ SAMPLE IDENTIFICATION DISTRICT UPAZILA UNION MOUZA VILLAGE TEAM CODE INTERVIEWER CODE Interview date (Month) 1=October; 2=November Interview date (Day of month) 122 1. Household Members (household head or spouse) Please tell the name of persons who usually live in your household (A household is a person or group of persons that usually lives and eat together and family members who lives outside visit the HH at least in every six months), starting with the head of the household. Table1: Household Members Line Is (NAME) male or female? How old is? (NAME) IF AGE LESS THAN 1 YEAR WRITE ‘00’ If aged 10 years or more: Educational Status If aged 10 years or more: Professions 101 102 103 104 105 01 M F 1 2 02 1 2 03 1 2 04 1 2 05 1 2 06 1 2 07 1 2 08 1 2 09 1 2 10 1 2 11 1 2 12 1 2 13 1 2 14 1 2 15 1 2 16 1 2 (Interviewer: Please note the line number of children <5) Are there any more household members (Yes/no) If yes, how many more members? _____- CODE LIST: Profession and Education Profession 01 = Do not work 02 = Household work 03 = Service 04 = Business 05 = Agriculture/ Farming 06 = Poultry 07 = Fish firming 08 = Daily wage earner 09 = Teacher 10 = Private Tutor 11 = Rickshaw/Van/Boat man/Driver 12 = Carpenter 13 = Weaver 14 = Cattle rearing 15 = Fisherman 16 = Tailor 17 = Others (Specify Education 01 = Ilterate 02= Can sign 03= Primary 04= Under SSC 05= SSC/Dhakhil 06= HSC/Alim 07= Bahelor/Fazil 08= Masters/Kamil 09= Others 123 2. Household Background Information (household head) NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 201 Main material of the roof Record observation No roof ................................. 1 Bamboo/thatch/palm leaf ...... 2 Tin ......................................... 3 Concrete ................................ 4 other ...................................... 5 202 Main material of the walls No wall .................................. 1 Bamboo/thatch/palm leaf ...... 2 Tin ......................................... 3 Wood ..................................... 4 Concrete ................................ 5 Other...................................... 6 203 How many does your household have of the following: (no=0) An Almirah (wardrobe)? A table chair or bench? A watch or clock? A cot or bed? A radio that is working? A television that is working? A bicycle? A motorcycle? A phone? A rickshaw/van/hand barrow/boat? 204 How much does your HH earn in a month (total of husband, wife and others)? (approximately) Monthly earning ( 205 How much did your HH earn in the last year from the following sources (none=0, don’t know=99)? Agric Livestock Salary (GoB, Private, teacher .. regular services with monthly payment) Daily labor Small business Merchant (Large business) Remittance (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) 124 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP Pensions (all types) Begging Rental income Professional (doctor, lawyer) Help from relative Money lending Other (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) (Tk.) 206 What is your HHs monthly expenditure? (Approximately) House rent .. (Tk.) Food .......... (Tk.) Utilities (electricity, gas, water, telephone) ................... (Tk.) (Tk.) Education ... (Tk.) Transport .... (Tk.) Medical ...... (Tk.) Loan repayment (T Others ........ (Tk.) 207 Does your household have access to khash land Yes ........................................ 1 No ............................................ 2 209 208 What is Khash land used for ? Living house Garden Cultivable land Forest Rent out Other 209 Does your household have access to water bodies Yes ........................................ 1 No ............................................ 2 301 210 What are water bodies used for? Fish Rent out Other 125 3. Agriculture (household head) NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 301 Does your household own cultivable land(including homestead land)? Yes ........................................ 1 No ......................................... 2 2 316 302 How much cultivable land does your household own (including homestead land)? (Exact in decimals) 303 Did you have any agricultural production in the last year (including gardening)? Yes ........................................ 1 No ......................................... 2 2316 304 How did agricultural production change last year compared to the year before? Increased ............................. 1 Stayed the same ................... 2 Decreased ............................ 3 Mixed ................................... 4 Don’t know .......................... 5 1- 305 2- 307 3- 306 4307 5307 305 Reasons for increase (Multiple response) More land farmed ................ 1 Better growing conditions ... 2 Better seed ........................... 3 More inputs used (fertilizer, etc) ...................................... 4 Improved irrigation .............. 5 Response to higher prices .... 6 Improved knowledge and skills ..................................... 7 Support from NGOs ............ 8 Improved pest management . 9 other ................................... 10 Skip 307 306 Reasons for decrease (Multiple response) Less land farmed .................. 1 Bad growing conditions ....... 2 Poorer seed .......................... 3 Fewer inputs available ......... 4 Response to lower price ....... 5 Less irrigation ...................... 6 Natural disaster .................... 7 Death/illness of family member(s) ............................ 8 pests ..................................... 9 126 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP other ................................... 10 307 Did your household sell any agricultural crops in the last year? Yes ........................................ 1 No ......................................... 2 2 313 308 If yes, what was value of sales? (Tk.) 309 How has the value of agricultural sales of your household changed in the last 3 years? Increased ............................. 1 Stayed the same ................... 2 Decreased ............................ 3 Don’t know .......................... 4 1310 2312 3311 4312 310 Reasons for increased sales (Multiple response) Less consumption by household ............................. 1 Greater area farmed ............. 2 Improved irrigation .............. 3 Better seed varieties ............. 4 Higher market prices ........... 5 Better market access ............ 6 Sale through farmers group . 7 Improved/lower cost transportation ....................... 8 Improved knowledge and skills ..................................... 9 Improved pest management .... 10 Skip 312 311 Reasons for decreased sales (Multiple response) More consumption by household ............................. 1 Decreased area farmed ........ 2 Flood .................................... 3 Drought ................................ 4 Lower market prices ............ 5 Less access in the market .... 6 Unavailability/high cost transport ............................... 7 Lack of irrigation ................. 8 Lack/high price of quality seed ...................................... 9 Pests ................................... 10 other ................................... 11 127 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 312 To whom did you sell agricultural products in the last year? (Multiple response) Neighbor/relatives/individuals .............................................. 1 Local market ......................... 2 Trader ................................... 3 Itinerant buyers ..................... 4 Cooperative/farmer group .... 5 Local Broker ......................... 6 NGOs .................................... 7 Company............................... 8 Collection point .................... 9 Other ................................... 10 313 Which of the following agricultural practices do you apply on your farm/garden? (Multiple response) Animal manure ..................... 1 Compost ................................ 2 Crop rotation ......................... 3 Chemical fertilizer ................ 4 Biological/organic pest control ................................... 5 Mechanical pest control ........ 6 Chemical pest control ........... 7 Integrated pest management 8 Treadle pump/drip irrig/mobile pump ................. 9 314 Have you received agricultural inputs from any of the following? Multiple response Local Market ........................ 1 Itinerant Merchants ............... 2 NGOs .................................... 3 GOB ...................................... 4 Companies ............................ 5 Cooperative/farmer group .... 6 Village Development Committee ............................ 7 Neighbor/relatives/individuals .............................................. 8 Trained input retailers ........... 9 Other ................................... 10 None ................................... 11 315 Have you received any training or technical support related to agriculture/gardening from any of the following? GoB office (BADC, BARI) .. 1 NGO ..................................... 2 Seed company ....................... 3 Others (specify) .................... 4 128 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP (Multiple response) No training ............................ 5 316 Did you have any livestock production in the last year? Yes ........................................ 1 No ......................................... 2 2321 317 How many of each of the following types of animals do you currently have? 1. Cows/buffalos 2. Goats/sheep 3. Chickens/ducks 4. Geese 5. Pigeon Number 318 How did livestock production change last year compared to the year before? Increased ............................. 1 Stayed the same ................... 2 Decreased ............................ 3 Mixed ................................... 4 Don’t know .......................... 5 1319 2321 3320 4321 5321 319 Reasons for increase Acquired more animals ...... 1 Improved breeds .................. 2 Better feed ........................... 3 Less disease ......................... 4 Response to better price ....... 5 Improved knowledge ........... 6 Support from NGOs ............ 7 Vaccination ........................... 8 Other ..................................... 9 Skip 321 320 Reasons for decrease Death/disease of animals .... 1 Animal stolen/lost ................ 2 Loss of land ......................... 3 Response to lower prices ..... 4 Disaster ................................ 5 Lack/high cost of feeds ........ 6 Lack of vaccine ..................... 7 Other ..................................... 8 321 Did you have any Fish production in the last year? Yes ........................................ 1 No ......................................... 2 2401 322 How did Fish production change last year compared to the year before? Increased ............................. 1 Stayed the same ................... 2 Decreased ............................ 3 1323 2401 3324 129 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP Mixed ................................... 4 Don’t know .......................... 5 4401 5401 323 Reasons for increase Better varieties of fingerlings 1 Lower cost fingerlings .......... 2 Improved knowledge ............ 3 Response to higher price ...... 4 Improved access to market ... 5 Support from NGOs ............. 6 Improved access/lower cost of feed ....................................... 7 Increased access to water bodies .................................... 8 Less disease .......................... 9 More fingerlings ................. 10 other .................................... 11 Skip 401 324 Reasons for decrease Less Access/Higher cost fingerlings ............................. 1 Response to lower price ........ 2 Less access/higher cost of feed ....................................... 3 Less access to water bodies .. 4 More disease ......................... 5 Natural disaster ..................... 6 Lower quality fingerlings ..... 7 Other ..................................... 8 130 4. Natural Disaster Preparedness (household head NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 401 What was the most recent type of natural disaster experienced in this area? Cyclone .................................. 1 Flood ...................................... 2 Earthquake ............................. 3 River erosion .......................... 4 Other (specify) ....................... 5 No disaster ............................. 5 5416 402 How long after the disaster did return to your home and start normal life? days 403 Did anyone in your HH die in last disaster (SIDR)? Yes ......................................... 1 No ........................................... 2 404 Did you lose any of the following? [Multiple response] House Livestock Documents Productive assets Household items Cash/jewelries 405 Did you receive any early warning signal/message before the last natural disaster (you had in your area)? Yes ......................................... 1 No ........................................... 2 2408 406 How long before the disaster did you receive the warning signal message? hours 407 Who gave the early/signal message? [Multiple response] CPP volunteers ....................... 1 Radio ...................................... 2 Television ............................... 3 Union parishad ....................... 4 NGOs ..................................... 5 Mosque miking ...................... 6 Neighbor/relatives .................. 7 Other (Specify) ....................... 8 408 Did you move to another place to take shelter before the last natural disaster? Yes ......................................... 1 No ........................................... 2 1410 409 If no, why not? No shelter No space available in the shelter Shelter not functional Did not receive messages No transport 131 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP Did not want To protect home/assets Live in protected house Others 410 Where did you move to take shelter before the last natural disaster? (Bold type indicates disaster-proof shelters) ‘Pacca’ House (cement) ....... 1 ‘Kacha’ house ........................ 2 Cyclone or flood shelter ....... 3 Union parishad building ......... 4 School/institution building ..... 5 Boat ........................................ 6 Highways/ Embankment ........ 7 Raised hillock ......................... 8 Mosque/Temple/Church ........ 9 Other (SPECIFY) ................. 10 411 How long before the disaster did you move to the shelter? (if during the disaster, enter 0 hours) hours 412 How far and long did it take you to go to the shelter centre for disaster? How far ..................... km Long .... Hrs. Mins. 413 After the last natural disaster, did you receive any assistance? Yes ......................................... 1 No ........................................... 2 2416 414 What did you receive? (Multiple response) Food ....................................... 1 Water ...................................... 2 Clothing .................................. 3 Housing .................................. 4 Money .................................... 5 Medicine ................................ 6 HH utensils ............................. 7 Others __________________ 8 415 When did you receive food and water? Just after the cyclone .............. 1 After 1 days ............................ 2 After 2 days ............................ 3 After 3 days ............................ 4 More than 3 days .................... 5 416 Are you aware of any members of the community trained to help you during disaster? Yes ......................................... 1 No ........................................... 2 2418 417 Who are they? CPP volunteers ....................... 1 Union parishad 132 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP (Multiple response) chairman/member .................. 2 NGOs ..................................... 3 Teacher ................................... 4 Students .................................. 5 Village leaders ....................... 6 Village Development Committee .............................. 7 Union volunteers .................... 8 Other (specify) ....................... 9 418 Have you or any member of your HH received any disaster preparedness training? Yes ......................................... 1 No ........................................... 2 2420 419 Who provided the training? CPP volunteers ....................... 1 Union parishad chairman/member .................. 2 NGOs ..................................... 3 Teacher ................................... 4 Students .................................. 5 Village leaders ....................... 6 Village Development Committee .............................. 7 Other (specify) ....................... 8 420 What do you plan to with your household members in the event of a disaster (cyclone/flood)? Don’t know ............................ 1 Evacuation of vulnerable HH members ................................. 2 Visit shelter centers in normal time ........................................ 3 Identify safe shelter center ..... 4 Plan for dry food .................... 5 other ....................................... 6 No plan .................................. 7 421 What do you plan to do with your livestock if a disaster strikes? Don’t know ............................ 1 Identify safe shelter for livestock ................................. 2 Arrange feed for disaster ........ 3 Assign a person responsible ... 4 other ....................................... 5 No plan ................................... 6 422 How do you plan to protect your HH valuables/assets in case of disaster? Don’t know ............................ 1 Arrangements to store assets 133 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP safely ...................................... 2 Assign a person responsible ... 3 other ....................................... 4 No plan ................................... 5 134 5. Food Security (wife, caregiver) NO. QUESTIONS AND FILTERS CODING CATEGORIES 501 I would like to ask you about the types of foods that you or anyone else in your household ate yesterday during the day and at night: Any: 1. Cereals (rice, noodles, bread) 2. Roots/Tubers (cassava, potatoes, sweet potatoes, plantains) 3. Legumes/Pulses (beans, peas, groundnuts, cashews) 4. Dairy products (milk, yogurt, cheese) 5. Meat (beef, offal, Poultry, mutton) 6. Fish/seafood 7. Oils, fats, butter, Ghee 8. Sugar/honey 9. Fruits 10. Eggs 11. Vegetables 12. Others Cereals ................................... Roots/Tubers ......................... Legumes/Pulses ..................... Dairy products ....................... Meat/poultry/offal ................. Fish/Sea food ........................ Oils/fat ................................... Sugar/honey ........................... Fruits ..................................... Eggs ....................................... Vegetables ............................. Others ..................................... 502 What type of salt does your HH consume regularly? Packet salt (observe) ............... 1 Loose ....................................... 2 503 In the past 12 months, were there months in which you did not have enough food to meet your family’s needs? Yes ............................................. 1 No .............................................. 2 504 If yes, which were the months (in the past 12 months) in which you did not have enough food to meet your family’s needs? [Multiple response] January ...................................... 1 February .................................... 2 March ........................................ 3 April .......................................... 4 May ........................................... 5 June ........................................... 6 July ............................................ 7 August ....................................... 8 September .................................. 9 October .................................... 10 November ................................ 11 December ................................ 12 135 NO. QUESTIONS AND FILTERS CODING CATEGORIES 505 In the past four weeks did you worry that your household would not have enough food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 506 In the past 4 weeks were you or any household member not able to eat the kinds of foods you preferred because of a lack of resources? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 507 In the past 4 weeks did you or any household member have to eat a limited variety of foods due to a lack of resources? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 508 In the past 4 seeks did you or any household member have to eat some foods that you really did not want to eat because of lack or resources to obtain other kinds of food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 509 In the past 4 weeks did you our any household member have to eat a smaller meal than you felt you needed because there was not enough food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 510 In the past 4 weeks did you our any household member have to eat fewer meals in a day because there was not enough food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 136 NO. QUESTIONS AND FILTERS CODING CATEGORIES Often (> 10 times in past 4 weeks)........................................ 3 Never ......................................... 4 511 In the past 4 weeks, was there ever no food of any kind to eat because of lack of resources to get food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never .................................... 4 512 In the past 4 weeks did you or any household member go to sleep hungry because there was not enough food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never .................................... 4 513 In the past 4 weeks did you or any household member go a whole day and night without eating anything because there was not enough food? Rarely (once or twice in past 4 weeks)........................................ 1 Sometimes (3-10 times in past 4 weeks)........................................ 2 Often (> 10 times in past 4 weeks)........................................ 3 Never .................................... 4 514 In the past 30 days, if there have been times when you did not have enough food or money to buy food, how often has your household had to (circle responses according to the scale below): 1 = Never 2 = Seldom (less than one day a week) 3 = Sometime (1-2 days a week) 4 = Often (3 or more days a week) 5 = Daily Limit portion size at meal times Reduce number of meals eaten per day? Borrow food or rely on help from friends or 1 2 3 4 5 1 2 3 4 5 1 2 3 4 137 NO. QUESTIONS AND FILTERS CODING CATEGORIES relatives? Rely on less expensive or less preferred foods? Purchase/borrow food on credit? Gather unusual types or amounts of wild food / hunt? Have household members eat at relatives or neighbors? Reduce adult consumption so children can eat? Rely on casual labor for food? Abnormal migration for work Skip entire day without eating Consume seed stock to be saved for next season 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 138 6. Safe Water, Sanitation, and Hygiene practices [Ask caregiver of children] NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 601 What are the main sources of water for drinking for your household? Deep tube well ...................... 1 Shallow tube well ................. 8 Pond sand filter ..................... 2 Rainwater harvesting system 3 Rainwater .............................. 4 Pond ...................................... 5 River/canal ............................ 6 Traditional well ..................... 7 Others (Specify) .................... 9 1,8602 2-7,9 604 602 Has the tube well you use been tested to see if its water has arsenic? [avoid of water source not tubewell, skip to water storage Tested .................................... 1 Not tested .............................. 2 Don’t know ........................... 3 2,3604 603 Is the tube well marked red or green? (Observe) Green ..................................... 1 Red ........................................ 2 Not marked ........................... 3 604 Do you store water in your home? Yes ........................................ 1 No ......................................... 2 2607 605 Do you collect and store drinking water in separate container? Yes ........................................ 1 No ......................................... 2 2607 606 Is the water kept covered? (observe) Yes ........................................ 1 No ......................................... 2 607 What type of latrine does your household use? (Bold type indicates hygienic types) Ring-slab/offset latrine (water seal) ....................................... 1 Pit latrine (covered) ............ 2 Ring-slab/offset latrine (water seal broken) ............... 3 Pit latrine (uncovered) .......... 4 Septic latrine ........................ 5 Hanging/open latrine ............ 6 No toilet facility .................... 7 6,7615 608 Is it your own latrine? Interviewer: Observe the latrine Yes ........................................ 1 No ......................................... 2 609 Do you use this latrine? Yes ........................................ 1 No ......................................... 2 610 When family members are at home, where do Male: Female: 139 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP your family members > age 5 go to defecate? Latrine ......... 1 Outside ........ 2 Latrine ......... 1 Outside ........ 2 611 Where do you dispose of your children’s feces? Latrine ................................... 1 Outside .................................. 2 Not applicable ...................... 3 612 Interviewer: Observe the following instruction 2.1.2 Is the latrine functioning? 2.1.2 Does the latrine show the sign of use? 2.1.2 Is the latrine (pan & slab) itself clean? 2.1.2 Is the surrounding area of the latrine clean? Latrine functioning ......... 1 ... 2 Shows the sign of use ..... 1 ... 2 Latrine itself clean .......... 1 ... 2 Surrounding area is clean 1 ... 2 613 When do you wash your hands? (Multiple response possible. DO NOT read the choices but probe and mark all that) Yes No Before food preparation .. 1 2 Before eating ................... 1 2 Before feeding children .. 1 ........................................ 2 After defication ............... 1 ........................................ 2 After cleaning babies bottoms ........................... 1 ........................................ 2 Others .............................. 1 ........................................ 2 (specify) 140 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 614 Can you show me how you wash your hands? (Go to hand washing site and observe hand washing technique that is demonstrated) (Multiple responses) Yes No Uses water ....................... 1 ........................................ 2 Soap/cleaning agent ........ 1 ........................................ 2 Ash .................................. 1 ........................................ 2 Washes both hands ......... 1 ........................................ 2 Rubs hands at least 3 times 1 Dries hands by air ........... 1 ........................................ 2 Dries hands with a clean cloth ................................ 1 ........................................ 2 Others (specify) .............. 1 ........................................ 2 Refused to demonstrate .. 1 ........................................ 2 7. Mothers/caregivers of children under 5 years: NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 701 Do you have any children under 24 months or are you currently pregnant? Yes ................................ 1 No .................................. 2 702 How old are you? (mother/care-giver of U5 in HH) Age (in completed years) 703 Did you ever attend school/madrasa? Yes ................................ 1 No .................................. 2 2706 704 Was it a primary school, madrasa, secondary school or higher that you attended last? Primary .......................... 1 Madrasa ........................ 2 Secondary School .......... 3 College/University ........ 4 Others ............................ 5 141 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP (Specify) 705 What was the highest class you passed? Class ................................ 706 Are you now married, widowed, divorced, or separated? Never married ............... 1 Currently married .......... 2 Widowed ....................... 3 Divorced ........................ 4 Separated ....................... 5 Deserted ........................ 6 707 Aside from doing normal household work, do you do any other work on a regular basis for which you are paid in cash or in kind or in both? Yes ................................ 1 No .................................. 2 2710 708 What do you do for your earning? (Multiple response) Handicrafts/Handloom .. 1 Agri/Farmer ................... 2 Work in other household3 Services ......................... 4 Business ........................ 5 Poultry ........................... 6 Daily wage earner ......... 7 Private tutor ................... 8 Others (Specify) ............ 9 No income earnings .... 10 709 How much do you generally earn a month from the activities you do? Monthly earning (Tk.) 710 Who usually makes decisions about how to to spend the cash income you earn? Husband ................................1 Wife .......................................2 Husband and wife jointly ......3 Somebody else ......................4 Husband and somebody else jointly 5 711 Who usually Husband ................................1 142 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP makes decisions about making major household purchases? Wife .......................................2 Husband and wife jointly ......3 Somebody else ......................4 Husband and somebody else jointly 5 712 Who usually makes decisions about purchases for daily household needs? Husband ................................1 Wife .......................................2 Husband and wife jointly ......3 Somebody else ......................4 Husband and somebody else jointly 5 713 Who usually makes decisions about visits to your family or relatives? Husband ................................1 Wife .......................................2 Husband and wife jointly ......3 Somebody else ......................4 Husband and somebody else jointly 5 714 Who usually makes decisions about your children’s health care? Husband ................................1 Wife .......................................2 Husband and wife jointly ......3 Somebody else ......................4 Husband and somebody else jointly 5 715 Are you currently pregnant? (Avoid if 701=2, skip to 800) Currently pregnant ................1 Not currently pregnant ..........2 Don’t know ...........................3 2,3712 716 How many months have you been pregnant for? Month(s) ....................... 717 Did you have any antenatal check￾ups during your (current/ last) pregnancy? Yes ........................................1 No ..........................................2 2717 718 How many check￾ups did you have during your (current/last) pregnancy? Number of visits ........... 719 Do you have an Yes, Seen ...............................1 143 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP antenatal card for your (current/last) pregnancy? IF Yes: May I see it please? Yes, Not Seen ........................2 No Card .................................3 720 Interviewer: Verify Number of Antenatal Visits Is the number of documented visits in the card different than the stated number of visits in Q204 Same as stated .......................1 Different than stated ..............2 Note number of documented visits ......... 721 Where did you receive ANC services? Hospital/Medical college .....1 Upazila Health Complex ......2 Satellite/EPI outreach centre 3 MCWC .................................4 FWC .....................................5 FWV .....................................6 FWA .....................................7 NGO Static clinic ..................8 NGO Satellite clinic .............9 NGO Field worker ..............10 NGO Hospital .....................11 VHC (village health committee CHV ....................................12 Clinic/Hospital ....................13 MBBS Doctor .....................14 Village doctor .....................15 Homeopathic doctor ............16 Pharmacy .............................17 Other Sector: Friend/Relative ....................18 Neighbor .............................19 Others (Specify) .................20 722 Did your husband Yes ........................................1 144 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP accompany you on any of your visits No ..........................................2 723 Did you receive Vita-A after delivery of the child? (Interviewer: shows her the Vit￾A capsule) Yes ........................................1 No ..........................................2 2719 724 After how many days of the delivery you received Vit-A? .............................. Days 725 Do you have a child of age <6 months? Yes ........................................1 No ..........................................2 145 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 726 Did you take yesterday or last night any of the following: 1. Milk/Dairy products (Calcium, VitA) 2. Roots/Tubers (potatoes, sweet potatoes, plantains) 3. Oils, fats and butter (VitA) 4. Fruits (Mango, Papaya, orange, Jackfruits)-VitA 5. Green leafy vegetables (VitA) Iron 6. Carrots/pumpkins (VitA) 7. Other fruit/vegetables 8. Egg (Vita) 9. Fish (Iron) 10. Poultry (iron) 11. Meat/offal/organs (iron) 12. Pulse/pea nuts/beans/ground nuts (iron) 13. Cereals Yes=1, No=2 2801 146 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 727 Have you taken Iron/Iron folate in the last 7 days? (Interviewer: show her the iron/iron folate tablet or capsule) Yes ........................................1 No……………………………………………………….2 8. Individual Child Related Questions NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 800 Do you have any children under 5? Yes .......................................... 1 No ........................................... 2 2901 801 Name of the youngest child 802 Sex of the youngest child Male=1, Female=2 803 Age of the youngest child Months 804 Did you ever breastfeed (NAME)? [avoid if U5 child code >1 and age >6 months, skip to 815] Yes .......................................... 1 No ........................................... 2 2812 805 How long after birth did you first put (NAME) to the breast? IF LESS THAN 1 HOUR, RECORD ‘00’ HOURS. HOURS ........................ 1 806 Did you give (NAME) the colostrum (the first milk which is yellow sticky fluid secreted the few days after delivery)? Yes .......................................... 1 No ........................................... 2 Don’t know ............................. 3 807 Did you give anything to (NAME) before the first breast milk? Yes .......................................... 1 No ........................................... 2 Don’t know ............................. 3 808 Did you give anything to (NAME) after starting breastfeeding? (Within 24 hours after starting breastfeeding) (Up to 3 responses allowed) No ........................................... 1 Milk (goat/cow/powder) ........ 2 Baby formula .......................... 3 Water/sugar water/honey ........ 4 medicine ................................. 5 809 Was (NAME) breastfed yesterday during the day or night? Yes .......................................... 1 No ........................................... 2 Don’t know ............................. 3 810 Did (NAME) have any of the following liquids yesterday during the day or night? Plain water .............................. 1 Sugar water ............................. 1 147 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP (Up to 5 responses allowed) Infant formula ......................... 1 Tinned, powdered, fresh animal milk ......................................... 1 Juice, juice drink, green coconut ................................... 1 yogurt ...................................... 1 ORS ........................................ 1 811 At any time yesterday or last night, was (NAME) given any liquid or solid food with breastfeeding? Yes .......................................... 1 No ........................................... 2 812 How many times yesterday or last night, was (NAME) given any of the following: (up to 10 responses allowed) 1. rice, bread, porridge, other foods made from grain 2. Pumpkin, carrots, orange sweet potatoes 3. White potatoes, white yams, other foods from roots 4. Dark green leafy vegetables 5. Ripe mangoes, ripe papayas, ripe jackfruits 6. any other fruits or vegetables 7. liver/kidney/heart 8. any meat (chicken, beef, goat, duck, etc.) 9. Eggs 10. Fish, dried fish, seafood 11. foods from beans, nuts, lentils 12. milk or milk products 13. oils, fats, butter, ghee 14. sugary foods such as chocolates, candies, pastries, cakes, biscuits 15. other 16. nothing Number of Times 148 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 813 Yesterday during the day or night, how many times did (NAME) eat solid, semi￾solid, or soft foods (foods other than liquids) at home or outside the home? (don't know =99) 814 Yesterday during the day or night did (NAME) drink anything from a bottle with a nipple? Yes ........................................ 1 No ......................................... 2 Don’t know ........................... 3 815 Yesterday during the day or night did (NAME) consume any food to which you added a nutrient powder (sprinkles/Monimix)? Yes ........................................ 1 No ......................................... 2 Don’t know ........................... 3 816 Did (NAME) receive a BCG vaccination against tuberculosis, that is, an injection in the left shoulder that caused a scar? YES ....................................... 1 NO ........................................ 2 DON’T KNOW .................... 3 817 Did (NAME) receive a polio vaccine that is, drops in the mouth? YES ....................................... 1 NO ........................................ 2 DON’T KNOW .................... 8 2,3819 818 How many times did (NAME) receive polio vaccine: From clinic? From NID? TIMES FROM CLINIC .... TIMES FROM NID ........... 819 Did (NAME) receive a DPT vaccination, that is, an injection given in the thigh or buttocks, sometimes at the same time as polio drops? YES ....................................... 1 NO ........................................ 2 DON’T KNOW .................... 8 2,3821 820 How many times? NUMBER OF TIMES ....... 821 An injection to prevent measles after 9 months of age? YES ....................................... 1 NO ........................................ 2 DON’T KNOW .................... 8 Not Applicable ...................... 8 822 Has (NAME) received a vitamin A capsule like this in the last 6 months? [avoid if age not 12-23 months, skip to diarrhea] Interviewer: Show Vitamin A Capsule Yes ........................................ 1 No ......................................... 2 Don’t know ........................... 3 823 Has (NAME) received antehelminth Yes ........................................ 1 149 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP (Deworming) within the last 6 months? [avoid if age not 12-23 months, skip to xxx] No ......................................... 2 Don’t know ............................. 3 824 Has (NAME) suffered from fever in the last 15 days? Yes ........................................ 1 No ......................................... 2 2 821 825 Did you seek advice/treatment for the fever of (NAME)? Yes ........................................ 1 No ......................................... 2 2  821 826 Where did you first seek treatment/advice for the fever of (NAME)? Hospital/Medical college ..... 1 Upazila Health Complex ..... 2 Satellite/EPI outreach centre 3 MCWC ................................ 4 FWC .................................... 5 FWV .................................... 6 FWA .................................... 7 Static clinic ........................... 8 Satellite clinic ...................... 9 Field worker ....................... 10 Hospital ............................... 11 CHV .................................... 12 Clinic/Hospital .................... 13 MBBS Doctor ..................... 14 Village doctor .................... 15 Homeopathic doctor ........... 16 Pharmacy ............................ 17 Friend/Relative ................... 18 Neighbor ............................ 19 Others (Specify) ................. 20 827 Has (NAME) suffered from cough/cold in the last 15 days? Yes ........................................ 1 No ......................................... 2 2 824 828 Did you seek advice/treatment for the cough/cold of (NAME)? Yes ........................................ 1 No ......................................... 2 2824 829 Where did you first seek treatment/advice for the cough/cold of (NAME)? Public Sector: Hospital/Medical college ..... 1 Upazila Health Complex ..... 2 Satellite/EPI outreach centre 3 MCWC ................................ 4 FWC .................................... 5 FWV .................................... 6 FWA .................................... 7 150 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP NGO Sector: Static clinic ........................... 8 Satellite clinic ...................... 9 Field worker ....................... 10 Hospital ............................... 11 CHV .................................... 12 Private medical sector: Clinic/Hospital .................... 13 MBBS Doctor ..................... 14 Village doctor .................... 15 Homeopathic doctor ........... 16 Pharmacy ............................ 17 Other Sector: Friend/Relative ................... 18 Neighbor ............................ 19 Others (Specify) ................. 20 830 Has (NAME) had diarrhea (having loose stool more than 2 times a day) in the last 2 weeks? Yes ........................................ 1 No ......................................... 2 Don’t know ............................. 3 2,3 END 831 Was (NAME) given the same amount to drink as before the diarrhea, or more, or less? Same ....................................... 1 More ....................................... 2 Less ......................................... 3 Don’t know ........................... 4 832 Was (NAME) given the same amount of food to eat as before the diarrhea, or more, or less? Same ....................................... 1 More ....................................... 2 Less ......................................... 3 Don’t know ........................... 4 833 Did you continue to breastfeed (NAME) during diarrhea? (avoid if 810 is no, skip to next question) Continued ............................... 1 Did not continue ................... 2 834 Did you seek advice or treatment for the diarrhea of (NAME) from any source? Yes ........................................ 1 No ......................................... 2 2830 835 Where did you first seek treatment/advice for the diarrhea of (NAME)? Public Sector: Hospital/Medical college ..... 1 Upazila Health Complex ..... 2 Satellite/EPI outreach centre 3 MCWC ................................ 4 FWC .................................... 5 151 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP FWV .................................... 6 FWA .................................... 7 NGO Sector: Static clinic ........................... 8 Satellite clinic ...................... 9 Field worker ....................... 10 Hospital ............................... 11 CHV .................................... 12 Private medical sector: Clinic/Hospital .................... 13 MBBS Doctor ..................... 14 Village doctor .................... 15 Homeopathic doctor ........... 16 Pharmacy ............................ 17 Other Sector: Friend/Relative ................... 18 Neighbor ............................ 19 Others (Specify) ................. 20 836 Did you give any of the following liquids/drinks to (NAME) for diarrhea in the last 15 days? (Multiple response) Fluid form ORS pkt .............. 1 Homemade sugar-water solution ................................. 2 salt-water solution (laban gur) .............................................. 3 Zink syrup ............................. 4 Zink tablet Fluid from special saline (rice) .............................................. 6 Nothing ................................. 7 9. Child rights and protection Questions NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 901 Do you know what are the rights of children in Bangladesh? non-discrimination (ethnic groups, disabled) ................................ 1 to live with parents ................. 2 to give opinion ....................... 3 to education ............................ 4 to health services .................... 5 to birth registration ................. 6 152 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP to recreation............................ 7 to protection from abusive child labor ....................................... 8 to protection from physical/social abuse....................................... 9 other ..................................... 10 don’t know 902 Do you agree or disagree with the following statement: It is wrong to hit children whenever they do something bad agree ....................................... 1 disagree .................................. 2 don’t know ............................. 3 903 What are the things that you believe children should be protected from? Physical abuse ........................ 1 Social stigma .......................... 2 Trafficking ............................. 3 Abusive child labor ................ 4 Early marriage ........................ 5 Sexual abuse ........................... 6 Physical/natural threats .......... 7 Other....................................... 8 Don’t know ............................ 9 153 9. Anthropometric Measurement: (separate form) SAMPLE IDENTIFICATION 1001 DISTRICT 1004 MOUZA VILLAGE 1005 TEAM CODE 1006 ANTHRO INTERVIEWER CODE 1007 HOUSEHOLD INTERVEWER CODE 1008 HH CODE (FROM HH INTERVIEW 1009 Interview date(month) October=1, November=2 1010 Interview date(day of month) NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 1011 Child Code (1=Youngest, 2=next to youngest, 3=next oldest) 1012 Child sex male ........................................ 1 female ..................................... 2 1013 Child birth date (YEAR) 2005 ........................................ 1 2006 ........................................ 2 2007 ........................................ 3 2008 ........................................ 4 2009 ........................................ 5 2010 ........................................ 6 Don’t know ............................. 7 1014 Child birth date (MONTH) Jan=1; Feb=2; Mar=3; Apr=4; May=5; Jun=6; Jul=7; Aug=8; Sep=9; Oct=10; Nov=11; Dec=12 1015 Child birth date (DAY) 1016 Child age in months (less than 1 month = 0) 1017 Child weight . kg 1018 Child length/height . CM 154 NO. QUESTIONS AND FILTERS CODING CATEGORIES SKIP 1019 Was length/height of child measured lying down or standing up? LYING.................................... 1 STANDING............................ 2 1020 Result CHILD MEASURED ............. 1 CHILD SICK .......................... 2 CHILD NOT PRESENT ........ 3 CHILD REFUSED ................. 4 MOTHER REFUSED ............ 5 OTHER ................................... 6 1021 Any more children U5? Yes ........................................ 1 No ......................................... 2 155 Annex 6: Additional Quantitative Data SO1 Tables Table 47. Overall and severe stunting, by age and district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall stunting (HAZ<‐2SD) children age 0‐59 months All households 43.9 31.8 ‐27.7 * 2,296   2,060   District         Barisal 50.0 37.3 ‐25.4 * 802    854        Barguna 37.7 25.9 ‐31.3 * 614    520        Patuakhali 42.8 29.5 ‐31.2 *   879    685   % of severe stunting (HAZ<‐3SD) children age 0‐59 months All households 12.9 9.0 ‐29.9 * 2,296    2,060   District         Barisal 17.7 11.3 ‐35.8 * 802   854        Barguna 9.8 6.5 ‐33.8 *   614   520        Patuakhali 10.7 8.1 ‐24.3 *         879   685   % of overall stunting (HAZ<‐2SD) children age 0‐23 months All households 33.2 19.1 ‐42.5 * 783    809   District         Barisal 35.3 22.6 ‐36.1 *      269    340        Barguna 27.9 14.9 ‐46.5 *    222    198        Patuakhali 35.3 17.7 ‐49.7 *      292   270   % of severe stunting (HAZ<‐3SD) children age 0‐23 months All households 9.0 5.8 ‐35.7 * 783   809   District         Barisal 8.5 8.3 ‐1.7 269   340        Barguna 11.0 4.6 ‐58.2 * 222   198        Patuakhali 8.1 3.6 ‐55.9 *     292    270   % of overall stunting (HAZ<‐2SD) children age 24‐59 months All households 49.5 40.0 ‐19.2 * 1,513   1,251   District         Barisal 57.4 47.0 ‐18.1 *   533   514        Barguna 43.2 32.6 ‐24.5 *   392   322        Patuakhali 46.5 37.1 ‐20.3 *   588   415   % of severe stunting (HAZ<‐3SD) children age 24‐59 months All households 14.9 11.1 ‐25.5 * 1,513   1,251   District         Barisal 22.3 13.4 ‐40.1 * 533   514        Barguna 9.1 7.6 ‐15.9   392   322        Patuakhali 12.0 11.0 ‐8.5    588   415   Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*).   156 Table 48: : Overall and severe stunting, by age and child sex   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall stunting (HAZ<‐2SD) children age 6‐59 months All households 43.9 35.3 ‐19.6 * 2,296 1,853 Child sex          Male 43.5 35.2 ‐19.1 * 1,136 944       Female 44.4 35.5 ‐20.1 * 1,160 909 % of overall stunting (HAZ<‐2SD) children age 6‐23 months All households 33.2 25.6 ‐22.8 * 783 602 Child sex          Male 32.7 27.9 ‐14.8 374 316       Female 33.6 23.1 ‐31.2 * 409 286 % of overall stunting (HAZ<‐2SD) children age 24‐59 months All households 49.5 40.0 ‐19.2 * 1,513 1,251 Child sex          Male 48.8 38.9 ‐20.3 * 762 628       Female 50.2 41.1 ‐18.1 * 751 623 % of severe stunting (HAZ<‐2SD) children age 6‐59 months All households 12.9 10.0 ‐22.2 * 2,296 1,853 Child sex          Male 13.2 11.3 ‐14.2 1,136 944       Female 12.6 8.7 ‐30.9 * 1,160 909 % of severe stunting (HAZ<‐2SD) children age 6‐23 months All households 9.0 7.8 ‐13.5 783 602 Child sex          Male 10.2 10.4 2.0 374 316       Female 8.0 4.9 ‐38.0 409 286 % of severe stunting (HAZ<‐2SD) children age 24‐59 months All households 14.9 11.1 ‐25.4 * 1,513 1,251 Child sex          Male 14.7 11.8 ‐19.7 762 628       Female 15.1 10.4 ‐31.1 * 751 623 Note:  Stars  indicate endline‐baseline difference is statistically significant at the 10% (*). 157 Table 49.  Overall and severe underweight, by age and district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall underweight (WAZ<‐2SD) children age 0‐59 months All households 39.4 27.3 ‐30.7 * 2,306 2,060   District         Barisal 40.1 27.8 ‐30.7 * 808  854        Barguna 37.4 25.8 ‐31.2 * 615 521        Patuakhali 40.1 27.8 ‐30.6 *   883 685   % of severe underweight (WAZ<‐3SD) children age 0‐59 months All households 9.9 5.2 ‐47.7 * 2,306 2,060   District         Barisal 11.4 5.4 ‐52.3 * 808 854        Barguna 7.9 5.0 ‐36.9 * 615 521        Patuakhali 9.8 4.9 ‐49.8 *     883  685   % of overall underweight (WAZ<‐2SD) children age 0‐23 months All households 31.9 19.5 ‐38.9 * 790 808   District         Barisal 30.7 21.1 ‐31.1 * 274 340        Barguna 29.0 16.3 ‐43.8 * 223 198        Patuakhali 35.2 19.7 ‐43.9 * 293 269   % of severe underweight (WAZ<‐3SD) children age 0‐23 months All households 7.6 4.5 ‐40.6 790 808   District         Barisal 6.5 5.2 ‐19.7    274 340        Barguna 5.0 4.8 ‐5.4      223 198        Patuakhali 10.6 3.5 ‐67.1 *         293 269   % of overall underweight (WAZ<‐2SD) children age 24‐59 months All households 43.3 32.3 ‐25.4 * 1,516 1,252   District         Barisal 44.9 32.1 ‐28.4 *    534 514        Barguna 42.2 31.6 ‐25.2 *    392 323        Patuakhali 42.6 33.1 ‐22.2 *   590 415   % of severe underweight (WAZ<‐3SD) children age 24‐59 months All households 11.1 5.6 ‐49.7 *   1,516 1,252   District         Barisal 13.9 5.6 ‐59.9 * 534 514        Barguna 9.6 5.1 ‐46.3 *        392 323        Patuakhali 9.5 5.9 ‐37.8 *        590 415   Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*).   158 Table 50: : Overall and severe underweight, by age and child sex   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall underweight (WAZ<‐2SD) children age 0‐59 months All households 39.4 27.3 ‐30.7 * 2,306 2,060 Child sex          Male 37.1 26.7 ‐28.0 * 1,142 1,046       Female 41.6 27.9 ‐33.0 * 1,164 1,014 % of overall underweight (WAZ<‐2SD) children age 0‐23 months All households 31.9 19.5 ‐38.9 * 790 808 Child sex          Male 30.0 20.5 ‐31.7 * 378 417       Female 33.6 18.4 ‐45.3 * 412 391 % of overall underweight (WAZ<‐2SD) children age 24‐59 months All households 43.2 32.3 ‐25.2 * 1,513 1,251 Child sex          Male 40.6 30.8 ‐24.0 * 762 628       Female 46.0 33.9 ‐26.4 * 751 623 % of severe underweight (WAZ<‐3SD) children age 0‐59 months All households 9.9 5.2 ‐47.7 * 2,306 2,060 Child sex          Male 8.7 4.7 ‐45.9 * 1142 1046       Female 11.0 5.6 ‐48.9 * 1164 1014 % of severe underweight (WAZ<‐3SD) children age 0‐23 months All households 7.6 4.5 ‐40.5 * 790 808 Child sex          Male 6.8 4.5 ‐33.1 378 417       Female 8.4 4.5 ‐45.9 * 412 391 % of severe underweight (WAZ<‐3SD) children age 24‐59 months All households 11.0 5.6 ‐49.7 * 1,516 1,252 Child sex          Male 9.7 4.8 ‐50.1 * 764 629       Female 12.4 6.3 ‐49.4 * 752 623 Note:  Stars  indicate endline‐baseline difference is statistically significant at the 10% (*). 159 Table 51.  Overall and severe wasting, by age and district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall wasting (WHZ<‐2SD) children age 6‐59 months All households 15.9 11.0 ‐31.0 * 2,296 1,851   District         Barisal 15.0 8.0 ‐46.6 *   803 768        Barguna 15.3 13.3 ‐12.9      613 477        Patuakhali 17.1 12.0 ‐29.7 *     879 606   % of severe wasting (WHZ<‐3SD) children age 6‐59 months All households 2.0 1.4 ‐30.5 2,296 1,851   District         Barisal 1.5 1.2 ‐17.6       803 768        Barguna 2.0 1.3 ‐33.3      613 477        Patuakhali 2.6 1.7 ‐32.4            879  606   % of overall wasting (WHZ<‐2SD) children age 6‐23 months All households 15.1 13.8 ‐8.8 780 600   District         Barisal 13.6 11.1 ‐18.2      269 254        Barguna 14.8 16.7 12.7      221 154        Patuakhali 16.7 14.9 ‐10.7      290 192   % of severe wasting (WHZ<‐3SD) children age 6‐23 months All households 3.0 3.0 ‐1.3 780 600   District         Barisal 1.1 2.7 143.6     269 254        Barguna 2.5 2.9 14.2    221 154        Patuakhali 5.2 3.5 ‐33.1           290 192   % of overall wasting (WHZ<‐2SD) children age 24‐59 months All households 16.3 9.6 ‐40.9 * 1,516 1,251   District         Barisal 15.8 6.5 ‐58.8 *    534 514        Barguna 15.5 11.7 ‐24.7      392 323        Patuakhali 17.3 11.9 ‐31.2 *     590 414   % of severe wasting (WHZ<‐3SD) children age 24‐59 months All households 1.5 0.7 ‐56.6 * 1,516 1,251   District         Barisal 1.7 0.5 ‐70.7 *       534 514        Barguna 1.7 0.6 ‐65.9      392 323        Patuakhali 1.3 0.9 ‐26.8     590 414   Note:  Stars for "all households" indicate endline‐baseline difference is statistically significant at the 10% (*).   160 Table 52: : Overall and severe wasting, by age and child sex   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of overall wasting (WHZ<‐2SD) children age 6‐59 months All households 15.9 11.0 ‐31.0 * 2,296 1,851 Child sex          Male 16.7 10.2 ‐39.2 * 1,135 943       Female 15.0 11.8 ‐21.8 * 1,161 908 % of overall wasting (WHZ<‐2SD) children age 6‐23 months All households 15.1 13.8 ‐8.8 780 600 Child sex          Male 15.4 12.2 ‐20.9 371 314       Female 14.8 15.5 4.7 409 286 % of overall wasting (WHZ<‐2SD) children age 24‐59 months All households 16.3 9.6 ‐41.0 * 1,516 1,251 Child sex          Male 17.4 9.2 ‐47.2 * 764 629       Female 15.2 10.0 ‐33.7 * 752 622 % of severe wasting (WHZ<‐3SD) children age 6‐59 months All households 2.0 1.4 ‐30.4 2,296 1,851 Child sex          Male 2.5 1.5 ‐39.0 1,135 943       Female 1.6 1.3 ‐17.6 1,161 908 % of severe wasting (WHZ<‐3SD) children age 6‐23 months All households 3.0 3.0 ‐1.2 780 600 Child sex          Male 2.8 3.5 26.4 371 314       Female 3.3 2.4 ‐25.4 409 286 % of severe wasting (WHZ<‐3SD) children age 24‐59 months All households 1.5 0.7 ‐56.7 1,513 1,250 Child sex          Male 2.3 0.5 ‐77.8 * 762 628       Female 0.7 0.8 16.7 751 622 Note:  Stars  indicate endline‐baseline difference is statistically significant at the 10% (*). 161 Table 53: Overall and severe child malnutrition indicators, by program participation   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline % of overall underweight (WAZ<‐2SD) children age 0‐59 months All households 39.4 27.3 ‐30.7 + 2,306 2,060 Program participation          Did not participate SO1 or SO2 22.9 ‐42.0 364       Participated SO1 only 27.3 ‐30.6 1,146       Participated SO2 only 31.4 ‐20.4 32       Participated SO1 & SO2 30.0 ‐23.8 518 % of severely underweight (WAZ<‐3SD) children age 0‐59 months All households 9.9 5.2 ‐47.7 + 2,306 2,060 Program participation          Did not participate SO1 or SO2 3.7 ‐62.6 364       Participated SO1 only 5.8 ‐40.8 1,146       Participated SO2 only 13.8 39.6 * 32       Participated SO1 & SO2 4.2 ‐57.9 518 % of overall wasted (WHZ<‐2SD) children age 6‐59 months All households 15.9 11.0 ‐31.0 + 2,296 1,851 Program participation          Did not participate SO1 or SO2 10.6 ‐33.4 307       Participated SO1 only 11.7 ‐26.5 1,036       Participated SO2 only 7.6 ‐52.1 28       Participated SO1 & SO2 9.9 ‐37.9 480 % of severely wasted (WHZ<‐3SD) children age 6‐59 months All households 2.0 1.4 ‐30.4 2,296 1,851 Program participation          Did not participate SO1 or SO2 0.6 ‐69.9 307       Participated SO1 only 1.9 ‐8.0 1,036       Participated SO2 only 4.7 130.6 28       Participated SO1 & SO2 0.8 ‐62.8 480 Note:  Plus sign (+) for "all households" indicate endline‐baseline difference is statistically significant at the 10%. Stars (*) for program participation indicate difference is statistically significant compared to "did not receive SO1 or SO2" at endline. 162 Table 54: Breastfeeding practices, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)    Number of observations    Baseline Endline Children under 6 month exclusively breastfed (%) All households 38.4 44.9 16.9 282 320 District         Barisal 29.1 39.3 35.2 * 133 151      Barguna 51.7 43.0 ‐16.8 54 75      Patuakhali 43.8 55.5 26.7 95 94 Children under 7 month exclusively breastfed (%) All households 34.2 42.5 24.3 * 323 348 District         Barisal 26.7 38.0 42.2 * 145 163      Barguna 47.0 40.6 ‐13.5 62 84      Patuakhali 36.7 51.3 39.9 117 101 Infants and toddlers who were put to the breast within one hour of birth (%) All households 28.9 41.1 42.1 * 1,142 968 District         Barisal 23.3 35.7 53.5 * 417 435      Barguna 26.5 52.3 97.6 * 297 241      Patuakhali 36.1 39.7 10.1 428 292 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 55: Incidence of and source of treatment for child diarrhea, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline % of children under 5 with diarrhea in last 15 days All households 10.4 7.5 ‐27.9 * 2,379 2,124 District         Barisal 12.3 9.9 ‐19.8 821 870      Barguna 8.4 7.9 ‐6.5 634 566      Patuakhali 10.1 4.2 ‐58.1 * 925 688 % of afflicted children who sought treatment All households 73.4 73.0 ‐0.5 248 160 District         Barisal 76.5 71.1 ‐7.1 101 86      Barguna 63.2 69.9 10.7 53 45      Patuakhali 75.9 83.5 10.1 93 29 163 Table 55: Incidence of and source of treatment for child diarrhea, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 56: Children with fever during the last two weeks, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline % of children under 5 with fever in last 15 days All households 54.9 48.8 ‐11.2 * 2,348 2,190 District         Barisal 56.4 49.3 ‐12.7 * 816 907      Barguna 51.8 50.6 ‐2.3 627 580      Patuakhali 55.8 46.6 ‐16.3 * 904 703 % of afflicted children who sought treatment All households 64.9 75.7 16.7 * 1,493 1,068 District         Barisal 70.4 73.7 4.6 549 447      Barguna 63.0 74.2 17.6 * 362 293      Patuakhali 60.8 79.8 31.3 * 582 328 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 57: Incidence of and source of treatment for child cough/cold, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baselin e Endline % of children under 5 with cough/cold in last 15 days All households 54.9 56.3 2.5 2,348 2,190 District         Barisal 56.4 58.6 3.9 816 907      Barguna 51.8 54.9 6.1 627 580      Patuakhali 55.8 54.4 ‐2.4 904 703 % of afflicted children who sought treatment All households 65.9 69.5 5.4 * 1,294 1,233 District         Barisal 70.9 68.1 ‐3.9 463 532 164 Table 57: Incidence of and source of treatment for child cough/cold, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)         Number of observations       Baselin e Endline      Barguna 63.8 68.0 6.5 326 319      Patuakhali 62.6 72.6 16.0 * 505 382 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 58: Child feeding and care giving practices, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline Infants/toddlers 6‐23  months who receive a minimally acceptable diet (apart from breast milk) All households 5.8 22.6 287.5 * 793 688 District         Barisal 6.1 20.6 236.9 * 260 301      Barguna 5.4 21.3 291.9 * 224 174      Patuakhali 5.9 26.5 349.8 * 310 214 Infants/toddlers older than 6 months who received iron rich/iron fortified foods during the previous day   All households 51.6 64.7 25.3 * 793 677 District         Barisal 43.1 54.7 26.8 * 260 296      Barguna 57.7 73.3 26.9 * 224 171      Patuakhali 54.3 71.8 32.2 * 310 210 Households consuming adequately iodized salt (20‐40ppm) All households 76.5 84.6 10.6 * 5,026 5,346 District         Barisal 67.4 81.9 21.6 * 1,649 2,031      Barguna 81.3 85.7 5.3 * 1,565 1,614      Patuakhali 80.6 86.6 7.5 * 1,812 1,701 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 59: Nutrient consumption among PLW, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Percentage of PLW who: 165 Table 59: Nutrient consumption among PLW, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Consume food rich in iron All households 31.5 90.5 187.1 * 431 517 District         Barisal 27.4 91.3 232.8 * 184 239      Barguna 24.3 90.5 272.3 * 100 122      Patuakhali 41.6 89.4 114.9 * 147 155 Consume food rich in vitamin A All households 22.3 59.6 167.0 * 431 517 District         Barisal 17.2 56.2 226.8 * 184 239      Barguna 28.0 70.7 152.0 * 100 122      Patuakhali 24.8 56.0 125.8 * 147 155 Consume food rich in calcium All households 12.3 12.4 1.1 420 517 District         Barisal 8.5 6.1 ‐27.8 124 168      Barguna 10.2 5.9 ‐42.8 162 165      Patuakhali 18.3 24.1 31.7 134 184 Have taken iron or iron folate supplements in the last 7 days All households 12.2 12.4 1.9 * 431 517 District         Barisal 10.2 13.5 32.1 * 184 239      Barguna 18.7 11.3 ‐39.6 * 100 122      Patuakhali 10.2 11.6 13.9 * 147 155 % of mothers of children aged 6‐23 months who received high‐dose Vitamin A supplement within 8 weeks postpartum (6 weeks if not exclusively breastfeeding) in last pregnancy All households 25.9 38.1 46.9 * 705 712 District         Barisal 16.4 37.6 129.7 * 218 306      Barguna 31.2 47.1 50.7 * 207 191      Patuakhali 29.5 30.9 4.8 280 215 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 166 Table 60: Attendance of antenatal care sessions, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of pregnant women or mothers of children under 2 attending at least 4 ANC sessions All households 11.8 32.9 178.1 *        1,145          1,095   District         Barisal 11.8 29.5 150.9 *            396              477        Barguna 10.1 41.8 315.3 *            316              282        Patuakhali 13.2 30.1 129.2 *            432              336   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 61: Caregiver hygiene practices, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)      Number of observations    Baseline Endline % of caregivers demonstrating proper personal hygiene behaviors All households 30.9 38.1 23.2 * 2,378 2,144 District         Barisal 27.7 39.1 41.4 * 820 887      Barguna 31.6 37.2 17.7 * 637 567      Patuakhali 33.3 37.4 12.3 * 921 691 % of caregivers demonstrating proper food hygiene behaviors All households 20.2 26.5 31.1 * 2,378 2,058 District         Barisal 19.4 28.3 45.3 * 820 848      Barguna 20.1 24.9 23.6 * 637 542      Patuakhali 21.0 25.6 22.0 * 921 667 % of caregivers demonstrating proper water hygiene behaviors All households 43.4 91.4 110.3 * 2,378 2,156 District         Barisal 47.9 86.8 81.0 * 820 893      Barguna 52.4 95.5 82.4 * 637 571      Patuakhali 33.3 93.9 182.0 * 921 692 % of caregivers demonstrating proper environmental hygiene behaviors All households 15.4 29.6 91.4 * 2,378 2,191 District         Barisal 16.8 27.8 65.5 * 820 907      Barguna 16.0 34.8 117.3 * 637 581      Patuakhali 13.9 27.5 98.5 * 921 703 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 167 Table 62: Percentage of children 12‐23 months who received Vitamin‐A supplementation, deworming treatment  w/in last 6 months, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of children that received Vitamin‐A supplementation All households 43.5 45.4 4.3            516              415   District         Barisal 44.9 48.4 7.9            165              178        Barguna 49.7 57.3 15.4            149              102        Patuakhali 37.8 32.4 ‐14.3            202              136   % of children 12‐23 months who received deworming w/in last 6 months All households 19.0 32.8 72.9 *            516              419   District         Barisal 17.9 35.4 98.3 *            166              182        Barguna 17.2 32.1 86.4 *            147              101        Patuakhali 21.2 29.8 41.0 *            203              136   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 63: Main source of drinking water, by district (endline, % of households)   Indicator Barisal Barguna Patuakhali Total Deep tube well 90.5 75.9 96.2 87.9 Shallow tube well 8.5 8.8 3.4 7.0 Pond 0.1 7.4 0.1 2.3 Pond sand filter 0.0 5.9 0.1 1.8 River/canal 1.0 0.1 0.2 0.5 Rainwater 0.0 0.9 0.0 0.3 Rainwater harvesting system 0.0 0.7 0.0 0.2 Other 0.0 0.3 0.1 0.1 N 2031 1614 1701 5346 168 Table 64: Main source of drinking water, by food security category (endline, % of households)   Indicator Low Middle High Total Deep tube well 87.7 86.7 89.4 87.9 Shallow tube well 6.6 7.4 6.8 6.9 Pond 2.3 2.7 1.9 2.3 Pond sand filter 2.0 1.9 1.4 1.8 River/canal 0.7 0.5 0.1 0.4 Rainwater 0.3 0.3 0.2 0.3 Rainwater harvesting system 0.2 0.3 0.2 0.2 Other 0.2 0.2 0.0 0.1 N 1787 1788 1760 5345 Table 65: Safety of tube well, by district   Indicator Barisal Barguna Patuakhali Total Mean of HH who use well as primary source of drinking water Tube well tested for arsenic 61.1 47.1 41.9 50.9 Not tested 13.7 20.9 26.3 19.8 Don't know 25.2 32.0 31.9 29.2 N 2010 1369 1694 5073 Status of testing:  Mean of HH where well was tested   Green 42.5 30.5 55.8 43.2 Red 0.9 1.9 1.8 1.4 Not marked 56.6 67.6 42.4 55.4 N 1228 645 709 2583 Table 66: Safety of tube well, by food security category   Indicator Low Medium High Total Mean of HH who use well as primary source of drinking water Tube well tested for arsenic 43.6 50.6 58.6 50.9 Not tested 23.0 18.8 17.5 19.7 Don't know 33.4 30.6 23.9 29.3 N 1685 1683 1693 5071 Status of testing:  Mean of HH where well was tested   Green 45.5 44.6 40.7 43.4 169   Indicator Low Medium High Total Red 2.1 1.6 0.9 1.4 Not marked 52.4 53.8 58.4 55.1 N 735 851 992 2582 Table 67: Water storage practices, by district (endline, % of households)   Indicator Barisal Barguna Patuakhali Total Store water in home 98.6 98.1 98.7 98.5 Drinking water is stored and collected in separate containers 90.4 95.1 94.5 93.1 Water is covered 86.6 93.5 93.1 90.7 N 2031 1614 1701 5346 Table 68: Water storage practices, by food security (endline, % of households)   Indicator Low Middle High Total Store water in home 98.6 98.3 98.5 98.5 Drinking water is stored and collected in separate containers 92.7 92.9 93.8 93.1 Water is covered 88.7 91.0 92.6 90.7 N 1778 1779 1779 5346 Table 69: Type of latrine, by district (endline, % of households)   Indicator Barisal Barguna Patuakhali Total Ring‐slab/offset latrine (water seal broken) 45.7 44.8 42.9 44.5 Ring‐slab/offset latrine (water seal) 31.7 30.4 31.4 31.2 Hanging/open latrine 2.8 9.0 10.3 7.0 Pit latrine (covered) 6.9 7.0 5.9 6.6 Pit latrine (uncovered) 6.7 6.0 5.9 6.2 Septic latrine 5.7 2.1 2.4 3.6 No toilet facility 0.6 0.8 1.2 0.8 N 2031 1614 1701 5346 170 Table 70: Type of latrine, by food security (endline, % of households)   Indicator Low Middle High Total Ring‐slab/offset latrine (water seal broken) 43.2 49.2 41.3 44.5 Ring‐slab/offset latrine (water seal) 24.1 28.4 41.2 31.2 Hanging/open latrine 12.6 5.6 2.9 7.0 Pit latrine (covered) 7.4 8.2 4.2 6.6 Pit latrine (uncovered) 10.1 5.5 3.1 6.2 Septic latrine 0.7 2.6 7.3 3.6 No toilet facility 2.0 0.6 0.0 0.8 N 1778 1779 1779 5346 171 SO2 Tables Table 71: Economic and food access indicators, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Number of income sources All households 2.2 2.6 17.3 *        5,026          5,346   District         Barisal 2.1 2.5 16.8 *        1,649          2,019        Barguna 2.2 2.5 15.0 *        1,565          1,615        Patuakhali 2.3 2.8 20.9 *        1,812          1,712   Average value of agricultural product  sales (Taka) All households 10448 11652 11.5 *        5,026          5,341   District         Barisal 8287 8692 4.9        1,649          2,017        Barguna 10668 11134 4.4        1,565          1,612        Patuakhali 12225 15628 27.8 *        1,812          1,712   Household Dietary Diversity Score (HDDS) All households 4.7 5.7 20.8 *        5,026          5,346   District         Barisal 5.0 6.1 22.2 *        1,649          2,031        Barguna 4.5 5.3 20.1 *        1,565          1,614        Patuakhali 4.7 5.5 18.3 *        1,812          1,701   Months of Adequate Household Food Provisions (MAHFP) All households 9.4 10.4 10.2 *        5,026          5,346   District         Barisal 9.8 10.4 6.0 *        1,649          2,031        Barguna 8.5 10.3 21.5 *        1,565          1,614        Patuakhali 9.9 10.4 5.6 *        1,812          1,701   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*). Sales values are reported as real, deflated values. Table 72: Household income and expenditures (in Tk), by food security category   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline) Number of observations (deflated) Baseline Endline Monthly Income Per Capita All households 1277 2345 1629 27.5 * 4,944 5,335 Food security category    172 Table 72: Household income and expenditures (in Tk), by food security category   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline) Number of observations (deflated) Baseline Endline       Low 897 1658 1152 28.4 * 1,648 1,787       Medium 1111 2137 1484 33.5 * 1,648 1,788       High 1824 3254 2260 23.9 * 1,647 1,760 Monthly Expenditures Per Capita All households 1528 2488 1728 13.1 * 4,944 5,335 Food security category          Low 1122 2052 1425 27.0 * 1,648 1,787       Medium 1368 2102 1460 6.7 * 1,648 1,788       High 2093 3323 2308 10.2 * 1,647 1,760 Food Share (%) of Total Expenditures All households 62.2 54.2 ‐12.9 * 4,944 5,335 Food security category          Low 70.5 59.4 ‐15.7 * 1,648 1,787       Medium 63.0 57.3 ‐9.1 * 1,648 1,788       High 53.2 45.7 ‐13.9 * 1,647 1,760 Asset Index All households 250.7 315.3 25.7 * 4,944 5,335 Food security category          Low 130.0 191.8 47.6 * 1,648 1,787       Medium 209.3 276.2 32.0 * 1,648 1,788       High 413.0 480.3 16.3 * 1,647 1,760 Asset Index Per Capita All households 52.0 69.0 32.8 * 4,944 5,335 Food security category          Low 28.9 43.5 50.7 * 1,648 1,787       Medium 44.4 63.3 42.7 * 1,648 1,788       High 82.7 100.7 21.7 * 1,647 1,760 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 73: Household income and expenditures (in Tk), by sex of head of household   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline)    Number of observations (deflated)    Baseline Endline Monthly Income Per Capita All households 1274 2344 1628 27.8 * 5,026 5,338 Sex head of household    173 Table 73: Household income and expenditures (in Tk), by sex of head of household   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline)    Number of observations (deflated)    Baseline Endline      Male 1258 2332 1620 28.7 * 4,722 4,997      Female 1515 2514 1746 15.3 * 304 342 Monthly Expenditures Per Capita All households 1520 2486 1727 13.6 * 5,026 5,338 Sex head of household         Male 1527 2475 1719 12.6 * 4,722 4,997      Female 1413 2645 1837 30.0 * 304 342 Food Share (%) of Total Expenditures All households 62.3 54.2 ‐13.0 * 5,014 5,335 Sex head of household         Male 61.9 53.9 ‐13.0 * 4,711 4,994      Female 67.3 58.6 ‐13.0 * 303 341 Asset Index All households 249.9 315.2 26.1 * 5,026 5,338 Sex head of household         Male 254.1 321.3 26.4 * 4,722 4,997      Female 183.9 226.2 23.0 304 342 Asset Index Per Capita All households 51.8 69.0 33.2 * 5,026 5,338 Sex head of household         Male 51.9 68.3 31.7 * 4,722 4,997      Female 50.2 79.0 57.3 304 342 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 74: Household income and expenditures (in US$), by district, food security, sex of head of household   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline)   Number of observations (deflated) Baseline Endline Monthly Income Per Capita All households 19 34 24 27.8 *               5,026                 5,338   District         Barisal 18 35 24 35.9 *               1,649                 2,019        Barguna 18 32 22 22.3 *               1,565                 1,610        Patuakhali 19 35 24 24.9 *               1,812                 1,709   174 Table 74: Household income and expenditures (in US$), by district, food security, sex of head of household   Indicator Baseline Endline Endline Percent difference (Endline ‐  Baseline)   Number of observations (deflated) Baseline Endline Monthly Expenditures Per Capita All households 22 36 25 13.6 *               5,026                 5,338   District         Barisal 22 34 24 6.7 *               1,649                 2,019        Barguna 22 40 28 27.2 *               1,565                 1,610        Patuakhali 22 35 25 9.4 *               1,812                 1,709   Monthly Income Per Capita All households 19 34 24 27.5 *               4,944                 5,335   Food security category          Low 13 24 17 28.4 *               1,648                 1,787         Medium 16 31 22 33.5 *               1,648                 1,788         High 27 47 33 23.9 *               1,647                 1,760   Monthly Expenditures Per Capita All households 22 36 25 13.1 *               4,944                 5,335   Food security category          Low 16 30 21 27.0 *               1,648                 1,787         Medium 20 31 21 6.7 *               1,648                 1,788         High 30 48 34 10.2 *               1,647                 1,760   Monthly Income Per Capita All households 19 34 24 27.8 *               5,026                 5,338   Sex head of household         Male 18 34 24 28.7 *               4,722                 4,997        Female 22 37 25 15.3 *                   304                     342 Monthly Expenditures Per Capita All households 22 36 25 13.6 *               5,026                 5,338   Sex head of household         Male 22 36 25 12.6 *               4,722                 4,997        Female 21 38 27 30.0 *                   304                     342 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*). Jan. 1, 2010 exchange rate of 68.80 used (source: www.xe.com). Table 75: Use of improved agricultural techniques, by food security category % household reporting using technique (endline)   Indicator Low Middle High Total 175 Fertilizer 77.7 81.3 83.6 81.2 Chemical pest control 61.9 68.1 71.5 67.8 Compost 29.7 32.6 34.3 32.5 Animal manure 30.1 27.3 31.2 29.6 Biological pest control 13.2 10.2 8.9 10.5 Crop rotation 9.0 10.0 11.5 10.3 Integrated pest management 7.7 10.8 11.1 10.1 Mechanical pest control 3.4 2.9 4.0 3.5 Improved irrigation 5.5 1.8 1.2 2.5 N 806 1027 1232 3065 Table 76: Use of improved agricultural techniques, by district % household reporting using technique (endline)   Indicator Barisal Barguna Patuakhali Total Fertilizer 77.0 82.3 84.3 81.3 Chemical pest control 68.1 67.5 67.8 67.8 Compost 36.4 35.9 25.5 32.5 Animal manure 29.3 29.2 29.4 29.3 Crop rotation 11.1 7.3 12.8 10.4 Integrated pest management 14.0 8.7 8.4 10.3 Biological pest control 14.0 6.8 10.2 10.2 Mechanical pest control 5.3 1.9 3.3 3.5 Improved irrigation 3.8 1.1 2.8 2.5 N 965 1046 1060 3071 Table 77: Households receiving agricultural training, by district % households that received training (endline)   Indicator Barisal Barguna Patuakhali Total Nobo Jibon 70.4 51.8 65.5 61.5 GOB 24.7 31.5 32.7 30.2 NGO 7.9 23.6 14.2 16.2 Seed company 6.6 8.6 6.0 7.2 Other 1.4 1.4 0.9 1.2 N 169 255 241 665 Table 78: Types of buyers for agricultural product, by food security % household reporting using buyers (endline)   Indicator Low Middle High Total 176 Local market 81.5 80.2 76.1 78.5 Traders 16.7 23.3 32.3 26.4 Neighbors/relatives 32.4 21.0 22.5 23.9 Local broker 4.4 7.2 10.3 8.2 NGO 3.0 3.4 2.3 2.8 Itinerant buyer 2.0 1.6 2.6 2.1 Other (collection point, cooperative, sales company) 2.9 3.2 2.4 2.7 N 587 950 1475 3012 Table 79: Types of buyers for agricultural product, by district % household reporting using buyers (endline)   Indicator Barisal Barguna Patuakhali Total Local market 75.8 84.2 75.9 78.6 Traders 28.6 25.2 26.3 26.5 Neighbors/relatives 20.7 24.8 24.8 23.8 Local broker 3.3 7.0 12.5 8.3 Itinerant buyer 3.0 3.0 0.9 2.1 Collection point 0.1 0.7 2.4 1.3 Other (NGO, cooperative, sales company) 0.0 0.1 0.2 0.1 N 781 997 1231 3010 Table 80: Use of marketing practices, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % HH adopting improved marketing practices All households 0.1 0.6 517.7 *        5,026          5,346   District         Barisal 0.1 0.0 N/A        1,649          2,019        Barguna 0.1 0.4 567.2 *        1,565          1,615        Patuakhali 0.2 1.6 799.4 *        1,812          1,712   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 81: Use and source of agricultural inputs, by food security category % household reporting purchase or receipt of inputs (endline)   Indicator Low Middle High Total 177 Companies 82.3 84.5 87.3 85.1 Neighbor/relative/individual 66.2 71.1 74.6 71.2 NGOs 28.1 32.6 37.0 33.2 Local Markets 30.4 26.8 32.6 30.1 GOB 9.9 10.6 12.9 11.3 Coops/farmer groups 13.4 10.7 9.7 11.0 Trained input retailer 7.5 10.9 11.4 10.2 Itinerant merchant 7.9 9.9 10.1 9.5 VDC 2.9 2.1 4.1 3.1 Other 2.7 2.3 1.6 2.1 N 653 842 1009 2504 Table 82: Use and source of agricultural inputs, by district % household reporting purchase or receipt of inputs (endline)   Indicator Barisal Barguna Patuakhali Total Companies 84.3 84.0 87.1 85.1 Neighbor/relative/individual 72.5 69.5 71.4 71.1 NGOs 37.5 35.6 27.3 33.3 Local Markets 28.4 30.8 30.2 29.8 GOB 11.5 8.8 13.7 11.4 Coops/farmer groups 14.8 7.2 10.7 10.8 Trained input retailer 14.4 7.9 9.2 10.4 Itinerant merchant 5.3 12.5 10.4 9.5 VDC 3.5 2.2 3.6 3.1 Other 2.5 3.1 0.8 2.2 N 788 847 876 2511 Table 83: Household production , by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline % HH with agricultural production last year All households 40.7 57.4 41.2 * 4,944 5,336 District         Barisal 23.9 45.3 89.6 * 1,648 1,778      Barguna 40.1 57.7 44.0 * 1,648 1,779 178      Patuakhali 58.0 69.3 19.3 * 1,647 1,779 % reporting increased agricultural production All households 40.1 47.9 19.4 * 2,025 3,071 District         Barisal 38.7 35.7 ‐7.7 569 965      Barguna 37.0 47.0 27.0 * 685 1,046      Patuakhali 43.9 59.9 36.3 * 771 1,060 % HH with livestock All households 61.5 82.4 33.9 * 5,026 5,346 District         Barisal 46.7 74.3 59.2 * 1,649 2,019      Barguna 75.5 85.2 12.7 * 1,565 1,615      Patuakhali 62.9 89.2 41.8 * 1,812 1,712 % reporting increased livestock production All households 26.8 21.7 ‐18.9 * 3,092 4,403 District         Barisal 27.8 15.9 ‐42.8 * 770 1,499      Barguna 25.8 22.0 ‐14.7 * 1,182 1,375      Patuakhali 27.1 27.2 0.2 1,140 1,528 % HH with fish production All households 22.7 30.3 33.2 * 5,026 5,346 District         Barisal 19.1 13.4 ‐29.9 * 1,649 2,019      Barguna 20.5 40.8 98.9 * 1,565 1,615      Patuakhali 28.0 40.4 44.1 * 1,812 1,712 % reporting increased fish production   All households 15.3 22.6 47.9 * 1,143 1,620 District         Barisal 16.7 24.5 47.3 * 315 270      Barguna 13.7 23.5 71.6 * 321 659      Patuakhali 15.4 21.0 35.9 * 507 691 % HH engaged in at least one category (crops, livestock, fish) All households 75.8 88.5 16.9 * 5,026 5,346 District         Barisal 65.0 82.1 26.3 * 1,649 2,019      Barguna 84.4 92.2 9.1 * 1,565 1,615      Patuakhali 78.0 92.7 18.8 * 1,812 1,712 % reporting increased production in any category 179 All households 38.8 43.6 16.9 * 3,807 4,733 District         Barisal 36.1 32.5 26.3 * 1,072 1,656      Barguna 37.3 45.3 9.1 * 1,322 1,489      Patuakhali 42.2 53.6 18.8 * 1,414 1,588 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 84: Access to agricultural land and water, by district   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % HH with agricultural land All households 59.2 68.0 14.8 * 5,024 5,346 District         Barisal 52.9 62.3 17.8 * 1,648 2,031      Barguna 63.4 74.5 17.4 * 1,564 1,614      Patuakhali 61.2 68.5 12.0 * 1,812 1,701 Average land area (decimals) All households 88.0 99.5 13.0 * 2,970 3,616 District         Barisal 82.2 85.6 4.2 871 1,260      Barguna 87.6 97.6 11.5 * 991 1,195      Patuakhali 93.1 116.5 25.2 * 1,109 1,161 % HH with access to khash land All households 10.1 11.1 9.8 4,944 5,335 District         Barisal 11.1 16.3 46.4 * 1,648 1,787      Barguna 11.2 9.7 ‐13.5 1,648 1,788      Patuakhali 7.9 7.2 ‐9.6 1,647 1,760 % HH with access to water bodies All households 64.3 80.5 25.3 * 4,941 5,335 District         Barisal 63.4 77.0 21.4 * 1,647 1,787      Barguna 69.8 81.9 17.3 * 1,647 1,788      Patuakhali 59.5 82.6 38.8 * 1,646 1,760 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 85: Economic and food access indicators, by food security category, by sex of head of household (US $) 180   Indicator Baseline Endline Percent difference (Endline ‐  Baseline) Number of observations Baseline Endline Average value of agricultural product  sales (US$) All households 153 169 10.7 *               4,944                 5,333   Food security category          Low 57 74 29.1 *               1,648                 1,785         Medium 108 143 33.2 *               1,648                 1,787         High 294 292 ‐0.6               1,647                 1,760   Average value of agricultural product  sales (US$) All households 152 169 11.4 *               5,026                 5,334   Sex head of household         Male 157 176 12.3 *               4,722                 4,993        Female 70 64 ‐9.8                   304                    342   Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) SO3 Tables Table 86: Household preparedness and impact of recent disaster, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Households have a plan to protect  HH members, livestock, or assets in the event of a disaster   All households 46.1 37.0 ‐19.7 * 4,944 5,335 Food security category          Low 43.1 33.5 ‐22.3 * 1,648 1,787       Medium 48.0 37.9 ‐21.0 * 1,648 1,788       High 47.0 39.6 ‐15.8 * 1,647 1,760 Minimal asset loss in the event of a disaster   All households 3.8 3.8 ‐19.7 * 4,944 4,405 Food security category          Low 3.4 2.4 ‐22.3 * 1,648 1,519       Medium 4.7 4.1 ‐21.0 * 1,648 1,449       High 3.1 5.0 ‐15.8 * 1,647 1,438 Able to resume livelihood activities within 2 weeks following a natural disaster   All households 4.6 22.1 383.4 * 4,944 5,335 Food security category          Low 2.8 20.8 653.6 * 1,648 1,787 181 Table 86: Household preparedness and impact of recent disaster, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline       Medium 3.4 23.9 610.8 * 1,648 1,788       High 7.6 21.6 184.3 1,647 1,760 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 87: Households who have received disaster preparedness training, by food security category   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline Households who have received disaster preparedness training   All households 4.6 22.1 383.4 * 4,944 5,335 Food security category          Low 2.8 20.8 653.6 * 1,648 1,787       Medium 3.4 23.9 610.8 * 1,648 1,788       High 7.6 21.6 184.3 * 1,647 1,760 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) Table 88: Early warning for disasters, by food security   Indicator Baseline Endline Percent difference (Endline ‐  Baseline)   Number of observations Baseline Endline % of households who sought shelter within 12 hours of  last disaster   All households 24.7 29.5 19.4 * 4,944 5,149 Food security category          Low 27.2 39.1 44.1 * 1,648 1,725       Medium 27.0 28.6 6.3 1,648 1,720       High 20.1 20.7 3.3 1,647 1,704 % of households who received warning within 12 hours of the last disaster All households 37.1 47.8 29.0 * 4,944 5,149 Food security category          Low 30.9 42.9 38.9 * 1,648 1,725       Medium 37.5 49.4 31.7 * 1,648 1,720       High 42.9 51.3 19.7 * 1,647 1,704 Note:  Stars indicate endline‐baseline difference is statistically significant at the 10% (*) 182 Annex 7: Terms of Reference, Baseline SCHEDULE A Terms of Reference Design and Implementation of Baseline Study and Development of M&E Plan Nobo Jibon Program - FY 2010-2015 Title II Multi-Year Assistance Program (MYAP) Name of Consultant : TANGO International (Mark Wyman Langworthy) Approximate dates : 16 August, 2010 to 15 January, 2011 (45 working days) Location : Dhaka, Barisal and other areas to be determined Supervised by : John Meyer, Chief of Party INTRODUCTION: Save the Children USA (SC) is commissioning a baseline study of its Title II Multi-Year Assistance Program, called Nobo Jibon, that will be implemented in three districts of Barisal Division in Bangladesh in collaboration with six local partner NGOs and three international technical partners. These TORs provide background information and expectations for the design and oversight of a baseline study, planned as a critical part of the eventual evaluation of the program. INTRODUCTION: Save the Children USA (SC) is commissioning a baseline study of its Title II Multi‐Year Assistance Program, called Nobo Jibon, that will be implemented in three districts of Barisal Division in Bangladesh in collaboration with six local partner NGOs and three international technical partners. These TORs provide background information and expectations for the design and oversight of a baseline study, planned as a critical part of the eventual evaluation of the program. PROGRAM BACKGROUND: The Nobo Jibon program has been designed to reduce food insecurity and vulnerability for 191,000 direct beneficiary households, or nearly 1 million people, in nine upazilas of Barisal Division over five years. The program comprises three strategic objectives (SOs) which are aligned with Bangladesh’s national health and food security policies and USAID’s priorities for Bangladesh. The SOs are: SO1 - Mother and Child Health and Nutrition (MCHN) - Improved health and nutritional status of targeted households, particularly children < five years of age SO2 - Market-based Production and Income Generation - Poor and extremely poor households have increased productivity and purchasing power to improve access to food SO3 - Disaster Risk Reduction -Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters. 183 Significant integration/overlap, i.e. households participating in all three SOs, will help assure greater impact than would be expected if interventions were dispersed. In total, Nobo Jibon will reach more than 1,300 villages and approximately 89% of the total 419,247 households in the nine target upazilas, which are: Barisal District Patuakhali District Barguna District Barisal Sadar Dashmina Amtali Hizla Galachipa Barguna Sadar Mahendiganj Kalapara Patharghata SO1 seeks to change childcare behaviors, improve intra-household food allocation, and integrate MCHN services and messages with GoB and private institutions. Nobo Jibon will provide a food ration to households with vulnerable women or children, conditional upon participation in awareness and education sessions. Behavior change communication (BCC) messaging will improve nutrition awareness and behaviors, community-based care of childhood illnesses, and hygiene practices. SO1 beneficiaries would total approximately 187,000 households in 79 unions. SO2 seeks increased productivity and income to improve access to food for such households. An income generation strategy will enhance agricultural and aquaculture productivity and profitability. Nobo Jibon will organize household groups, help build technical skills for increased horticultural, fish, poultry or non-farm production and improve links to markets. The program will promote access to khas resources and improve sustainable access to capital to meet input/service needs. This component will target 80,000 poor and extremely poor households in all Nobo Jibon communities. An additional 9,000 extremely poor households will be targeted for asset transfers, to catalyze new income generating activities. Additional economic benefits, such as increased access to quality inputs and services; increased market activity; improved market infrastructure; and improved technologies may indirectly benefit an additional 100,000 beneficiary households. SO3 activities will directly or indirectly benefit all households (circa 373,470) within the core geographic area targeted by the program. All SO1 and SO2 beneficiaries will benefit from risk reduction, with 44 unions determined to be highly disaster prone targeted during a first phase. Food for work (FFW) and/or case for work will provide a safety net, while helping build DRR infrastructure. SC’s involvement in multi-agency disaster preparedness networks will extend some benefits of the program (e.g. advances in early warning systems) beyond the nine targeted upazilas. Given known vulnerabilities in upazilas elsewhere in the division, SC proposes that its emergency contingency planning and response activities consider the entire Barisal Division as its target. PURPOSE OF THE ASSIGNMENT: Consultant support is required for assisting in the development of an M&E plan leading to the overall design and management of a baseline survey, along with a thorough analysis of data and presentation of findings. The M&E plan to be finalized following the FANTA‐2 M&E workshop in August 2010 will lead to an appropriate baseline survey design. The baseline study aims, through a quantitative survey of a representative sample of households in the program impact area, to establish pre‐program benchmarks for key indicators, to help refine program targets and to help prioritize program activities. External consultant expertise is required 184 to assure appropriate sampling strategy and data collection methods and to objectively analyze, interpret and present data. STATEMENT OF WORK: A sequence of activities is proposed for this assignment. The following provides detail on specific tasks for the consultant(s). A. Participate in a M&E workshop offered to newly-awarded Title II Multi-Year Assistance Programs (MYAPs):  The consultant needs to join in the M&E workshop organized by FANTA-2 in August 16-20, 2010 in Bangladesh to have a better understanding on FFP new strategies in M&E. This participation will be essential in designing an M&E plan in line with FFP guidelines and priorities. B. Assist in developing Nobo Jibon M&E Plan and staff capacity:  Work with the Nobo Jibon M&E Manager to develop a comprehensive M&E plan for Nobo Jibon to be submitted to USAID for approval before the baseline.  Work with the Nobo Jibon M&E Manager to design and deliver an M&E workshop for key Nobo Jibon staff (SC and partners) following content of FANTA workshop. C. Develop Baseline Sampling Methodology, Survey Instruments, and Survey Design Document  Review Nobo Jibon program document and IPTT indicators and discuss information needs with key stakeholders.  Prepare draft questionnaire, solicit feedback, finalize questionnaire  In consultation with stakeholders, devise a sampling strategy that results in the collection of data required for fulfilling survey objectives, while economizing on time and resources.  Submit for approval a concise but comprehensive design document describing all steps in survey methodology, including the analyses proposed D. Program Software and Personal Digital Assistants (PDAs)  Using software of consultant’s choice, develop computer-based questionnaire template, assuring interface with PDAs, including application of Bengali fonts.  Put in place a system for data management, including uploading of data collected in appropriate form and format.  Train staff, as required, to manage the system F. Train Staff, including Field Supervisors, and Pre-test Instrument  Prepare and deliver six-day training including two days for field testing for a team of enumerators and field supervisors.  Lead a separate session with team leaders and quality control team to agree on systems for assuring the quality control of data collected.  Conduct and debrief a field pre-test of the survey instrument, making corrections to questions and methods as required G. Oversee Data Collection  Oversee first rounds of data collection and provide guidance and feedback to local team  Be available for remote problem solving in case of need. 185 H. Analyze Data and Present Results  Thoroughly analyze the data collected.  Disaggregate data and conduct comparative analysis as possible among geographic and/or demographic sub-groups.  Present findings in tabular and graphic format with narrative descriptions and interpretations. TEAM COMPOSITION: Local team members will be contracted outside of this contract to fulfill the requirements of the baseline study. A local team leader will help in designing methodology, tools and training data collectors. This local consultant will be recruited by Save the Children through a separate contract. Additional team members to be recruited locally by Save the Children will include a survey supervisor, field team leaders, quality controllers and enumerators. A sub‐set of enumerators will be experienced in anthropometric measurement. Save the Children M&E staff will provide assistance to resolve technical issues related to survey design, tools development, recruitment of the data collection team, training and field implementation process. APPROXIMATE TIMELINE AND LEVEL OF EFFORT: Prepare for and attend the FANTA‐2 M&E Workshop (06 days): The international consultant will join in the M&E workshop to newly‐awarded Title II Multi‐Year Assistance Programs (MYAPs) in Bangladesh. M&E planning (06 days) Remote contributions (drafting, reviewing, editing) the Nobo Jibon M&E plan and baseline study design. Conduct M&E Workshop for Nobo Jibon Staff (September 20-22: 03 days): Following content of the FANTA-2 M&E workshop the consultant will facilitate a M&E workshop for Nobo Jibon stakeholders. Design Quantitative Population-based Survey (September 26 – 30: 05 days): The survey team leader will review background documentation and conduct interviews with SC, and other stakeholders to develop quantitative survey instruments, sampling plan, while outlining a plan and time line for team recruitment and training, data collection and analysis. A survey design document will be produced. Form and Train Team of Enumerators (October 02 – October 07: 05 days): The survey team leader will prepare a team of enumerators and team-leaders adequate to collect the required data within the time allotted. S/he will design and deliver training, as needed. Pre-testing of data collection should be part of the training schedule. Coordinate Quantitative Population-based Survey (October 09-17: 09 days): The survey team leader will be present at the beginning of data collection only, assuring that appropriate data 186 collection and management methods are used and that the local field supervisors are able to lead the process. Data entry and cleaning will be conducted as needed using selected analysis software. Quantitative Data Analysis and Reporting: Data analysis and preliminary interpretation of the findings will be done by the survey team leader. S/he will prepare a survey report summarizing findings (approx 11 days for analysis, writing and review). A total of up to 45 working days will be made available to consultant for full implementation of these TORs. REPORTING AND DELIVERABLES: There are three written deliverables for this assignment:  A revised/edited Nobo Jibon M&E Plan by September 20, 2010.  A baseline study design document and work plan, finalized by September 30.  A draft quantitative survey summary report in English by December 15 with final version with dataset and syntax files by 10 January 2011. Section 6. Statement of Work (SoW) Final Statement of Work (SoW) Quantitative Performance Evaluation (QPE) of the Nobo Jibon Program USAID Supported Title-II Multi-Year Assistance Program (MYAP) Save the Children International in Bangladesh 1. INTRODUCTION Save the Children International (SCI) in Bangladesh is commissioning a Quantitative Performance Evaluation of the United States Agency for International Development (USAID) supported Title II Multi Year Assistance Program (MYAP), called Nobo Jibon, being implemented in Bangladesh where four Implementing Partners (GUP, Speed Trust, SAP Bangladesh and CODEC) implementing the whole program in the field. Nobo Jibon has technical partnership with four other organizations (HKI, IDE, WORLDFISH & RIMES) for technical backstopping from the beginning. This SoW describes the specific objectives, approach and expected deliverables for the Quantitative Performance Evaluation of Nobo Jibon. The goal of this evaluation is to assess both the impact of program activities as measured by positive differences within indicators of interest at baseline and end-line and the extent to which the measured differences achieve the Life of Award targets and to draw reasonable inferences of successful association between the program strategies and the impacts on the behavior and well-being of the beneficiary population. Program Food for Peace (Title-II) Project Title Nobo Jibon Awardee Save the Children Award# AID- FFP-A-10-0-00011 Award duration 1-Jun’2010 to 31-May’2015 Funding Amount $ 55.73 M (USAID, GoB & SC) Implementing Partners Community Development Center (CODEC) Gonno Unnuyan Prochesta (GUP) South Asian Partnership (SAP) Bangladesh Speed Trust Government Partner Department of Relief and Rehabilitation (DMRD) Technical Partners Helen Keller International (HKI) International Development Enterprises (IDE) World Fish Centre Regional Integrated Multi-Hazard Early Warning System (RIMES) AOR Shahnaz Zakaria CoP (in-charge) Bakaul Islam Email: bakaul.islam@savethechildren.org Address: Road 43, House-CWN (A) 35, Gulshan-2, Dhaka 1212 1.1 Description of the Nobo Jibon Program The Nobo Jibon program has been designed to reduce food insecurity and vulnerability for 191,000 1 direct beneficiary households, or nearly 1 million people, in eleven2 upazilas of Barisal Division over five years. The program comprises three strategic objectives (SOs) which are aligned with Bangladesh’s national health and food security policies and USAID’s priorities for Bangladesh. The SOs are: SO1 - Maternal Child Health and Nutrition (MCHN) - Improved health and nutritional status of Children under the age of five (U5) and Pregnant and Lactating Women (PLW) seeks to change childcare behaviors, improve intra-household food allocation, and integrate MCHN services and messages with GoB and private institutions. Nobo Jibon will provide a food ration to households with vulnerable women or children, conditional upon participation in awareness and education sessions. BCC messaging will improve nutrition awareness and behaviors, community-based care of childhood illnesses, and hygiene practices. Complementary water and sanitation interventions will support households of greatest need. SO1 beneficiaries would total approximately 225,000 households in 86 unions. Figure-1: Program Operational Area SO2 (Market-based Production and Income Generation): Poor and extremely poor households have increased production and income to improve access to food. SO2 seeks increased productivity and income to improve access to food for poor and extremely poor households. A comprehensive income generation strategy is applied to enhance agricultural, aquaculture, small scale livestock & non-farm productivity, profitability and competitiveness in the market. Livelihood component has organized household groups, build their technical skills for increased production and improved market linkages in the above subsectors. The program promoted access to khas 3 resources and piloted 320 Village Savings and Loan Associations (VSLA) as model for accessing finance. Activities were undertaken to make the livelihoods poor beneficiaries resilient to natural shocks. This component targeted 80,000 poor and extremely poor households in Nobo Jibon communities. Interventions with the input and output market actors and service providers, improved market infrastructure, and improved technologies may indirectly benefit an additional 100,000 beneficiary households. SO3 - Disaster Risk Reduction - Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters. Figure-2: Program Integration (10%) activities will directly or indirectly benefit a majority of households (approximately 444,241) within the core geographic area targeted by the program. All SO1 and SO2 beneficiaries will benefit from risk reduction, with 53 unions determined to be highly disaster SO1 186,930 SO2 (90%) 72,000 8,000 SO3 373,470 prone targeted during a first phase. Food-for-work will provide a safety net, while helping build DRR infrastructure and livelihood resilience. Structural intervention (new and rehabilitation of cyclone 1 Nobo Jibon proposal document figure is 191,000 based on design level estimation; however, based on program operation trend, management decided to extend the figure to 225,000 and thus got approval through last PREP submission. 2 During FY 2012 Galachipa was divided into two upazila’s of Galachipa & Rangabali. Similarly, in Amtali) was divided into two upazila’s of Amtali & Taltali. While the geographic coverage has not changed, the program now works in 11 upazilas of Barisal Division for the remaining period. 3 Khasland or state-owned land is the land which the government is entitled to both lease and give away to citizens of the country who do not own land. Strategic Objectives Nobo Jibon Interventions and Outputs SO1: Maternal Child Health and Nutritio n: Improved health and nutritional status of Children under the age of five (U5) and • Formed and trained 1156 Village Health Committees on community-based management and treatment of childhood illness and malnutrition • Provided training on Essential Nutrition Actions at community and government levels, Village doctor and TBA. • Community-based growth promotion (with MoHFW) for approximate SO2: Market-based Production and Income Generation: Poor and extremely poor • Training and demonstrations on contextually appropriate technologies and practices to 40,000 Homestead Productive Poor (HPP) (vegetables/fish) who are also accessing quality inputs and market information • Asset transferred to at least 16,535 out of 20000 registered Extreme Poor (EP) to catalyze income generation shelters) reduced the vulnerability of affected community. SC’s involvement in multi-agency disaster preparedness networks will extend some benefits of the program (e.g. advances in early warning systems) beyond the nine targeted upazilas. Given known vulnerabilities in upazilas elsewhere in the division, the Nobo Jibon will be responsible for emergency contingency planning and response activities throughout the Barisal Division. Significant overlap, i.e. households participating in all three SOs, will help assure greatest impact (graphically depicted in Figure 1). In total, Nobo Jibon will reach more than 1,300 villages and approximately 89% of the total 444,241 households in the eleven target upazilas (see the map below). There will be significant overlap for greatest impact that will happen when households participate in all three SOs, as diagrammed in the left side. M&E and Information Management A comprehensive monitoring and evaluation (M&E) plan has been developed with detailed descriptions on baseline, annual, semi-annual, quarterly and monthly monitoring to track the program progress and outcomes over the period of time. The Nobo Jibon IPTT (initially approved by the AOTR in October 2010) is used by the program for annual reporting to USAID. The IPTT includes baseline figures, collected and compiled after the original IPTT was approved. The targets of the IPTT were revised based on the baseline figures and the experience of the team over Implementation Year. Annual monitoring has been done during September-October period since 2011 to meet the Annual Results Report (ARR) of USAID and other reporting purpose. Semi-Annual Monitoring (SAM) is designed to measure the longitudinal effect of the program intervention in production and income generation. Each year, two rounds (March & September) of this type of monitoring is done and have planned to complete six rounds starting from March’2012. The progress of the program activities is tracked monthly using the custom designed McAID system. All SO1, SO2 and FFW beneficiaries are registered through the McAID system, which tracks their activities as well. Tools and techniques have been developed to apply qualitative monitoring system to collect success stories quarterly and longitudinal case studies to track progress and describe impact at the beneficiary level. M&E staffs are maintaining the qualitative monitoring system regularly and have created a data bank for success stories and case studies to share with different stakeholders. Table 1: SOs with corresponding interventions/outputs are summarized below: Strategic Objectives Nobo Jibon Interventions and Outputs administration and mobilization” of whom 77 have accessed to khash land SO3: Disaster Risk Reduction: Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters. • Target Households: 444,241 • 739 high risk and Risk villages trained and equipped for Disaster Risk Reduction (DRR) (e.g. maps, plans, volunteers) • 7 new cyclone shelters constructed; 71 cyclone shelters rehabilitated • FFW/LRA undertaken in 550 communities resulting in nearly 1.835 million person days of employment • Pilot early warning to track depression and landfall for enhanced accuracy, improved lead time and community level application Save the Children Quantitative Performance Evaluation (QPE) SOW 71 2. QUANTITATIVE PERFORMANCE EVALUATION SCOPE OF WORK The main objective of the Quantitative Performance Evaluation (QPE) is to review a) the achievements of the project relative to its prescribed targets and b) progress towards the overall goal of positive impact on food security of target communities. 2.1 Evaluation purpose The purpose of the quantitative performance evaluation is to evaluate the performance of key indicators against the baseline values to measure strategic objectives and intermediate results of Nobo Jibon. Specific objectives include: • Assessing whether progress against agreed indicators/targets have met end of project benchmarks as documented in the indicator tracking table; • Evaluating the theory of change through establishing plausible links between inputs, outputs, outcomes and impacts on target population; • Determine whether critical strategies are missing that were needed to achieve Nobo Jibon’s goal; • Assessing the overall impact of the project on target population; • Identify where interventions, in isolation or in combination, were insufficient to meet program goals and, in cases where goals were not met, assess whether that was due to faulty logical reasoning/hypothesized causal pathways, to implementation shortcomings, or to other factors ; and, • Suggest design adjustments to improve the quality of future programming. 2.2 Evaluation scope In order to achieve these objectives the Consulting Team / Contractor will be expected to: 1. Design and implement a population-based quantitative survey that captures necessary information needed to achieve the evaluation objectives. This will require careful consideration and documentation of potential and selected sampling frames and justification of the selected approach. 2. The Consulting team/ Contractor must follow the appropriate Performance Indicator Reference Sheets developed by Food for Peace and Feed the Future, to finalize the survey instrument and methodology. In designing the instruments, the Contractor must consider including all disaggregate levels required by USAID/ FFP. 3. The survey instrument used in the baseline survey should be the starting point. If there is a need for rephrasing any question, the Contractor must add a question with revised statement instead of changing an existing one. Changing a question will void the comparability. In consultation with Nobo Jibon staff and the USAID/Bangladesh Mission, the Contractor may add questions to the survey instrument. 4. A Survey Plan must be prepared and get approved by USAID before the survey implementation. The design document should include sampling strategy and sample size estimation, sampling frame and household listing, data treatment and analysis plan, training of enumerators and supervisors, field testing of the instruments, and oversight and quality control mechanisms. The Contractor must specify the details of the sampling design in the survey design document in advance of field implementation. This document must include all of the following elements: i. The principal indicator and associated target group that will drive the sample size calculation for the entire survey. ii. The Contractor should show the equation used for this calculation and the parameters used in the equation, including the design effect assumed for the principal indicator driving the sample size calculation. The calculation should take into account statistical power. iii. The number of households to be sampled in order to achieve the desired sample size for the target group (assuming that households may contain more than one or no eligible members from the target group). The Contractor should give an indication of how the base sample size will be adjusted to account for the number of households that need to be visited. Design effect should be used from the baseline survey data. iv. The number of households to be sampled to account for anticipated household non￾response. The Contractor should indicate by how much the number of households to be sampled will be pre-inflated to account for household non-response. v. Geographic or other criteria for stratification. The Contractor should specify all stratification criteria and the total number of strata for all criteria. vi. The number of stages of sampling to be used. vii. Explanation of how the number of clusters and of households per cluster in the sample will be determined. viii.Explanation of how individual members in a sample household will be selected for interview (this is particularly important for anthropometric indicators, questions related to women at reproductive age, questions related to infant and young child feeding practices and farmers) ix. Definition of the clusters. The Contractor should use tables to show the number of clusters that will be selected for each stratum. x. Explanation on the source of information for the sampling frame, e.g. census lists or other national or internationally-sponsored surveys, such as the Demographic Health Surveys (DHS) or household listing. If secondary data are used, the Contractor should indicate how reliable and recent the frame information is. xi. A Probability Proportionate to Size (PPS) sampling mechanism should be used to randomly select the clusters. The Contractor should use the number of households per cluster as the size measure and include a table of size measure and another showing the final list of selected clusters along with their probabilities of selection. xii. Indication that the Contractor will use systematic sampling to select dwellings within clusters. This implies that for the sampled clusters, a list of all households, with household identification and location indicated, within these clusters must be obtained through either a preliminary pass on the cluster prior to interviewing or other existing sources. 5. Data Treatment and Analysis Plan: The Contractor must prepare a data treatment and analysis plan to address the following elements: i. Indication of how and when data will be entered into the database, as well as the software to be used for data entry. Double-data entry is required; If smart phone, PDA, or tablet is used to capture data, name of the application and the strategy to double-check the data on a regular basis so that any inconsistencies can be identified immediately and corrective measures can be taken within a day; ii. Data quality checks and edits (data cleaning) planned to ensure logical consistency and coherence, as well as an indication of the software to be used; 6. Sampling weights to be included on the data file. The formulae used to calculate the sampling weights should be included as part of a data dictionary document. Different sampling weights will need to be calculated for separate analysis of each district and of the program level aggregate. Note that a household non-response adjustment should be made to the sampling weights as part of the final weighting system; 7. Indicator tabulation plan. Estimates should be produced for each stratum and for the overall level; Indication of which sub-groups, if any, for which the Contractor will produced estimators; i. To understand factors that explain the variation in change in stunting, household hunger scale, household dietary diversity score, and minimum acceptable diet, multivariate analysis model must be specified and presented in the tabulation plan. ii. The contractor should specify all intended bivariate and multivariate analysis in the tabulation plan; iii. Indication that confidence intervals associated with the indicators will be produced alongside the indicator estimates and that these will take into account the design effect associated with the complex sampling design. Additional statistical outputs are required for multivariate analysis, but should be provided in an appendix; iv. Software to be used for data analysis and for conversion of anthropometric data into Z-scores. v. Description of methods for comparing the final survey data with the baseline survey data, and tests to be used to detect a population level difference at 95 percent level of significance. 8. The tools and methodologies should be finalized having concurrence from Nobo Jibon management and USAID/ FFP, OFDHA/B. 9. Field Procedure Manual: It is expected that the Contractor will develop a field manual to be used as part of the training materials for survey enumerators and supervisors and serve as reference material for staff in the field conducting the survey. The field manual should include instructions on how to sample dwellings within clusters, households within dwellings, and select individuals within households. The manual should also give recommended best practices for conducting interviews and dealing with specific challenging situations, e.g. households that refuse to participate, and provide a household and individual respondent non-response follow-up strategy. The manual should also describe the roles and responsibilities of the enumerators, supervisors, and other field staff and contain a detailed explanation of how to properly administer each question in the questionnaire. 10. The survey team is required to ensure statistically representative data collection within the household survey and anthropometric measurement with appropriate representation of three districts. 11. For the anthropometric data collection, the Contractor must use international standard height boards and weight scales. 12. To comply with USAID’s Open Data Policy, USAID/FFP will host the data to USAID’s Open Data portal. To comply, the Contractor must submit the following: i. Raw data and the cleaned data files with all of the computed variables both in SPSS and CSV formats. ii. SPSS or STATA Syntax files and weighting files in Microsoft Excel iii. Submit a data dictionary - essentially a definition and description of any of the fields provided in the dataset 13. Carry out a comparative analysis of quantitative survey results, existing M&E data and/or the data presented in Indicator Performance Tracking Tables (IPTT) as needed 14. Establish plausible links between inputs, outputs, outcomes and impact at final evaluation and identify factors that contribute to or impede the program interventions. 15. Quantify program result with comparison to Baseline and disaggregated by district, gender and economic strata and document accordingly. 16. Find reasons of program progress and challenges encountered to produce targeted results (if any). 17. Make specific recommendations based on results of the quantitative survey on how the project could have improved its strategies, systems and interventions to enhance its performance with respect to the above mentioned objectives. 18. Consideration/assessment of beneficiary targeting criteria and practice, including transparency, probing a.) whether the right beneficiaries are identified for program interventions and b.) whether the interventions are appropriate for the most vulnerable beneficiaries (and if not, why not). There should also be some consideration of barriers to participation for vulnerable groups (childcare availability, travel distances, gender norms of women in public, etc.). 2.3 Audience and use of information The primary audience of the evaluation report will be USAID (Bangladesh and Washington DC), and specifically the Office of Food for Peace. The report will provide important evidence to Government of Bangladesh (line ministries) and Save the Children (Bangladesh, Member & SCI) and other MYAP Partners. Secondary users might be development organizations (PVOs and Development actors), other interested donors (UN, DFID, EC, etc.), and research/academic institutions as a contribution to the broader. 2.4 Indicators to evaluate SL Level Indicator description 1 GOAL % children between 6 and 59 months stunted (height-for-age) (disaggregated by gender) (<- 2SD; <-3SD) 2 GOAL Average HH Food Insecurity Access Scale score 3 GOAL Average HH coping strategy index 4 SO1 Percentage of underweight (WAZ<-2) children aged 0-59 months 6 SO1 Percentage of wasted (WHZ<-2) children aged 6-59 months 8 SO1 % Children 0-6 months exclusively breastfed 10 SO1 % of children 6-23 months of age who receive a minimum acceptable diet (apart from SL Level Indicator description breastmilk) 12 SO1 % of caregivers demonstrating proper personal hygiene behaviors 13 SO1 % of beneficiary caregivers demonstrating food hygiene behaviors 14 SO1 % of PLW who consume food rich in iron 16 SO1 % of PLW who consume food rich in Vitamin A 18 SO1 % of PLW who consume food rich in Calcium 20 SO1 % of PLW taking iron or iron folate supplements in the last 7 day 22 SO1 % of children 12-23 months who received Vitamin-A supplementation in the past 6 months 23 SO1 % of mothers of children aged 6-23 months who received high-dose Vitamin A supplement within 8 weeks postpartum (6 weeks if not exclusively breastfeeding) in last pregnancy 24 SO1 % of mothers attended ANC session at least 4 times during last pregnancy 25 SO1 % of beneficiary children 12-24 months receiving de-worming medication in previous 6 months 30 SO1 % of beneficiary women whose husband attends ANC/PNC with her 31 SO2 Average HH dietary diversity score (HDDS) 32 SO2 Average number of months of adequate household food provisioning (MAHFP) 33 SO2 % of HHs reporting increase in production of one or more products 34 SO2 Average annual incomefrom sale of agricultural products 35 SO2 Per capita yearly income of USG targeted beneficiaries 39 SO2 %of beneficiaries (farmers) using 3 or more sustainable/improved production practices. 41 SO2 Numbers of farmers and others who have applied new technologies or management practices as result of USG assistance 43 SO2 Number of hectares under improved technologies or management practices as a result of USG assistance. 44 SO2 % of targeted PP HHs adopting improved marketing practices 48 SO2 % of targeted HHs (PP+HPP) having access to quality inputs and technical service 49 SO2 % of targeted HHs (PP+HPP) having access to or participating in output markets 53 SO2 % of extremely poor HHs using distributed assets for increased production and income generation. 54 SO3 % of HHs with a feasible plan to protect human life and productive assets during disaster 55 SO3 %of HHs able to resume livelihood activities within 2 weeks following a natural disaster. 63 SO3 % of HHs that sought shelter in a timely manner during last disaster 64 SO3 % of HHs that received location specific cyclone warning signal with adequate lead time 2.5 Evaluation Questions Through the analysis of the quantitative final survey data, most recent monitoring indicator results, Midterm Review data and supportive qualitative assessments with key stakeholders, the contractors will address the following questions: MATERNAL CHILD HEALTH AND NUTRITION (SO1) 1. To what level Nobo Jibon achieved the anthropometric targets stated at IPTT? 2. What level were MCHN and WASH practices adopted by the community in comparison with Baseline, Geography, Sex, HH economic strata? 3. Do the NJ livelihood supported activities have positive correlation with higher uptake of MCHN service and behavior? If so, what beneficiary segments and/or interventions show maximum correlation? 4. Is there any quantitative evidence suggesting the hypothesis that PLW and mothers with children under 2 that participated in MCHN sessions from Nobo Jibon are more likely to practice MCHN and IYCF practices and contribute to better nutritional outcomes? 5. Whether ration size difference had an effect on MCHN practice/Outcome controlling other factors? MARKET BASED PRODUCTION & INCOME GENERATION (SO2) 1. To what extent were proposed SO2 program targets achieved? 2. Were NJ mechanisms/techniques adopted to make livelihoods resilient to natural disasters? 3. What percent of extreme poor HHs practice effective and sustainable IGA through NJ support and which IGAs proved to be sustainable and suitable for those? 4. Which market-driven approaches taken by Nobo Jibon contributed to the improvement of the productivity and income of the targeted beneficiaries (PP and HPP)? 5. Did introduced technologies have significant contributions to the food security of HH? Which categories of HHs achieved success and what contributing factors were responsible? 6. What technologies are mostly adopted by HHs? What factors (geography, HH economic category, land type, program follow-up etc.) are causing the adoption rate? DISASTER RISK REDUCTION (SO3) 1. To what extent were proposed program output targets were achieved? 2. Was the HH level awareness program successful in gaining knowledge and practice of DRR initiatives taken by HH members? 3. Are vulnerable areas people more aware in terms DRR preparedness? 4. How successful are the assisted communities in achieving the preparedness of HHs? 5. What is the level of uptakes of NJ promoted messages by HHs? 6. How far communities are familiar with cyclone signaling system and use that during NJ period? CROSS CUTTING (COMMUNITY MOBILIZATION & GENDER) 1. To what extent community committee (VDC) contributed in Nobo Jibon program achievement and sustainability of intervention? 2. To what extent has the project improved role of women in: • Decision making about MCHN and care? • Decision making about intra-household food distribution especially for PLW and Children under 2? • Decisions about production system, income, savings and usage of resources? • Decisions about making major household purchases? SUSTAINABILITY 1. To what extent, Nobo Jibon beneficiaries are habituated in taking services from local level government, service delivery sources and NGOs? 2. What are the HHs’ perception about the quality and effectiveness of services provided by NJ and perceived source of services after NJ ends? QUANTITATIVE PERFORMANCE EVALUATION DESIGN & METHODOLOGY 2.6 Evaluation design This assignment will be a Quantitative performance evaluation which is mainly quantitative survey but accompany with some qualitative research methods to satisfy the validation of quantitative data for concrete result reporting. The quantitative survey will utilize the same ‘adequacy design’, or non￾experimental design for simple pre-post comparison of results as Baseline design. This will be a population-based survey with the sample drawn randomly from the sample frame of all households residing within the action areas of Nobo Jibon. The sample size will be determined to provide statistically representative results for indicators at the level of household and children under five years of age. A two-stage sample selection process will be used to select households to be interviewed. In the first stage, mouzas will be selected in each of the three program districts. In the second stage, a pre-determined number of households will be interviewed in each of the selected mouzas to achieve desired confidence intervals. Before selecting the HHs, A complete HH census will be done from each selected Mauzas by the enumerators. Enumerators will go from door to door, compiling a list of all households in the mauza with name, location, information on selection criteria for being selected as evaluation respondent. Then, from that list a pre-determined number of eligible households will be selected randomly. During analysis the sample will be weighted to account for the fact that within the three districts, the proportion of sampled households to district population is different and non￾response cases should be counted during sample size determination process. As per USAID policy and Nobo Jibon M&E, Quantitative Performance Evaluation sampling process should be same as Baseline protocol. To be noted that in Baseline, total sample size was 5082 where per district sample was 1694. During Quantitative Performance Evaluation sampling, P1 and design effect value should be used from Baseline data. On the other hand, non-response factor should be used from the experience of Baseline and Midterm review. While a majority of the data will be collected to measure progress against indicators in the IPTT, data for some indicators will be drawn from the project M&E system and the beneficiary based annual survey. In addition, to satisfy the interpretation of Quantitative data and reporting, some qualitative tools will be applied in addition to the population based survey. These may include key stakeholder informant interviews and/or focus group discussions. For example, Nobo Jibon is working mainly with farmers under Livelihood component but there are some interventions with non-farmers stakeholders like market players (seed dealers, input suppliers, buyers etc.). So, contribution to market system should be captured by interviewing both farmers and other market players including the GO￾NGO service providers. While USAID will commission a separate and independent qualitative evaluation, the Quantitative Performance Evaluation will measure changes in performance indicators to determine the performance of Nobo Jibon and its contribution towards national and USAID’s strategic goal. This study will complement the qualitative evaluation report. 2.7 Data Collection Method The Quantitative Performance Evaluation data will be collected through a Structured questionnaire from population within the selected Nobo Jibon community. Others stakeholder information through either quantitative or qualitative tools should be collected as per the evaluation design and SoW questions. In addition to the quantitative population based survey the consultants may use qualitative interviews and other tools to gather information from private sectors (seed dealer, paikars, agents, buyers, company representatives, government staffs, NGOs etc. They should consider employing a variety of qualitative primary data collection methods, including semi-structured in-depth-interviews, group discussions, key informant interviews, and direct observation. 2.8 Data Analysis Method Data analysis will be done with several relevant comparisons among various groups. An analysis plan should be finalized and shared prior to data exploration. The analysis may present comparisons of food security levels among identified groups using factor analysis. (Refer to the Baseline report and M&E Plan). Additionally, data will be disaggregated and compared by the three sampled districts: Barisal, Barguna, and Patuakhali. Last, for several key food and economic security indicators, the data is analyzed by sex of household head. Data should be presented with statistical significant tests and confidence intervals whenever it will be compared with Baseline figure. Consultant team will analyze household data using statistical software (SPSS/STATA etc.), calculating secondary variables (asset indices, coping strategy index) from primary variables where appropriate. Where appropriate, data must be weighted to account for differences in underlying population distributions among the primary sampling units. To validate the quantitative data, evaluation team should interview different stakeholders through different qualitative tools like FGD, KII, semi-structured interview, case studies etc. and analyzed the context holistically for preparing the Quantitative Performance Evaluation report. 3 QUANTITATIVE PERFORMANCE EVALUATION PRODUCTS 3.1 Deliverables 1. The following deliverables are to be submitted to Save the Children: The Consultant team/ Contractor is responsible for: a) Pertinent permissions, insurance, and other required permits i). Obtaining all the necessary permissions for implementing the quantitative evaluation data collection. ii). Adhering to country and local formalities and obtaining any required permits related to data collection from human subjects and logistics of survey implementation, including any necessary Internal Review Board (IRB) approvals, as well as health and accident insurance, salary, and taxes for all enumerators and supervisors. Deliverable: Evidence of insurances and permits for implementing survey and other data collection activities in electronic form b) Survey plan including detailed survey implementation plan (DSIP) i). Specifying details for methodology, critical tasks, anticipated outputs, date-bound timelines, resource needs, and responsible person(s). Composition of a standard field survey team, including expected tasks and responsibilities of each team member, should also be described. ii). Detailing a sampling plan for the quantitative population-based household survey that responds to the elements specified in Section 2.2. Deliverable: Survey plan including sampling plan, and detailed implementation plan reviewed and approved by Save the Children, USAID/FFP, USAID/Bangladesh Mission. c) Quantitative survey instrument which must take into account the instrument used in the baseline survey. Additional questions can be added to the instrument if needed. Adapting the questionnaire to the local context if additional questions are to be added to the instrument. Translating the approved questionnaire instrument from English into Bangla. Back translating the questionnaire from Bangla to English with a second translator to ensure it is accurately translated in Bangla. Making any necessary changes to Bangla questionnaire based on the back translation. The questionnaire used in the baseline survey is already translated into Bangla, thus the contractor does not have to translate the major part of the questionnaire. Deliverable: Final Bangla and corresponding English questionnaires reviewed and approved by Save the Children, USAID/FFP and USAID/ Bangladesh Mission d) Data treatment and analysis plan a. Detailing a data treatment and analysis plan that responds to the elements specified in section 2.2.4. Deliverable: Data treatment and analysis plan reviewed and approved by Save the Children, USAID/FFP and USAID/Bangladesh Mission e) Raw and cleaned data set, data dictionary/codebook, edit rules, and syntax for data analysis, including syntax for variable transformations Deliverables: i. Raw data set in SPSS and CSV formats;: ii). Edit rules for cleaning data; iii). Data dictionary/codebook; iv). Syntax for all data analysis and variable transformations; v). Final data set that includes cleaned data, sampling weights at each stage, final sampling weights, and all derived indicators; and vi). Sampling weights used to tabulate the aggregate-level estimates for the USAID/FFP Standard Indicators f) Briefings for the Save the Children and USAID Bangladesh. Presenting findings, conclusions, lessons learned, and recommendations based on the quantitative performance evaluation survey. Deliverables: i). Mid-term and final briefings to Save the Children and final debriefings to USAID Bangladesh g) Draft quantitative performance evaluation report i). Not exceeding 50 pages, excluding appendices and attachments. The draft report must be presented in English. ii). Presenting the estimates and confidence interval for all indicators (impact and outcome) at the Nobo Jibon program level and by districts; iii). Using appropriate tests of differences, determine the change at the underlying population level with confidence intervals. Deliverable: Draft quantitative performance evaluation report reviewed and approved by Save the Children and USAID h) Final Quantitative Performance Evaluation report: This report will be a revised version of the quantitative performance evaluation report that incorporates the comments of Save the Children, USAID/FFP and the USAID Bangladesh Mission. The final report must be presented in English. Deliverable: Final quantitative performance evaluation report reviewed and approved by Save the Children and USAID Reporting guidelines Final Quantitative Performance Evaluation reporting should be done based on USAID’s Evaluation policy4 and other guideline on reporting structure5 . The report should be a maximum of 50 pages, in Times New Roman 12pt font in single space, excluding the cover page, table of contents, and annexes. The format for the Quantitative Performance Evaluation report is as follows: 4 This policy can be accessed at http://www.usaid.gov/evaluation/USAID_EVALUATION_POLICY.pdf. 5 Available at http://www.usaid.gov/policy/evalweb/documents/TIPS-ConstructinganEvaluationReport.pdf 1. Cover page, Table of Contents, List of Acronyms 2. Executive Summary should be a clear and concise stand-alone document that states the most salient findings, conclusions, and recommendations of the evaluation survey and gives readers the essential contents of the Quantitative Performance Evaluation survey report in two or three pages. The Executive Summary helps readers to build a mental framework for organizing and understanding the detailed information within the report; 3. Introduction should include purpose, audience, and synopsis of task; 4. Program background should describe Nobo Jibon goal, theory of change, targeting, geographic scope, history and key timeline 5. Methodology should describe sampling design, study methods, data collection techniques, constraints and limitations of the study process and rigor, and issues in carrying out the study; 6. Tabular summary of results should present quantitative performance evaluation results in table form for all the indicators by district and for the overall program; 7. Findings should present findings on all of the key indicators. Quantitative performance evaluation survey values must be presented in quantitative format and complemented by descriptive analysis for each stratum and at the aggregate Nobo Jibon program level; 8. Conclusions and Recommendations should provide additional analysis of the data and results, drawing out programmatic and organizational recommendations for future program design. Recommendations must be relevant to program and context. 9. Issues should provide a list of key technical and/or administrative, if any, for Nobo Jibon for which the quantitative performance evaluation survey was conducted; and 10. Annexes should document the study methods, scope of work, schedules, interview lists and tables and be succinct, pertinent, and readable. 11. References, including bibliographical documentation, meetings, interviews, and focus group discussions; 12. List of stakeholder group with number, type, and date of interactions; 13. Data collection instruments in English and the local language; 14. Data dictionary and program files used to process the data in electronic format; 15. Other special documentation identified as necessary or useful. 16. Data Tables showing progress against the IPTT + SAPQ indicators (by district). 17. Tabular results with Statistical significance i. Data analysis & Statistical test syntaxes 3.2 Deliverables & Timeline Output Jun’2014 Jul’2014 Aug’2014 Sep’2014 Oct’2014 Nov’2014 Dec’2014 Jan’2015 1-15 16-30 1-15 16-31 1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-31 Final SoW submission to USAID accommodating the comments Approval from USAID RFP floated Selection of Consultant firm QPE team meeting with NJ and other SC teams Work Plan finalization Draft FE protocol, including detailed SoWs for each expert for the team, as well as the team leader Questionnaire design, translation, data collector training materials, field guides preparation, Design data entry templates at appropriate hardware (PDA/Smartphone) as per finalized Survey questionnaire Arrange necessary survey equipment (PDA/Smartphone, Weight scale, height scale, weight, stationeries etc.) Enumerators training, field testing the questionnaire, data-entry templates, data entry hardware and measuring equipment etc. Conduct Quantitative performance evaluation survey as per FE protocol Stakeholder meetings and other information (non-population based data collection) collection for data validation. Submit draft FE report. Submission of final QPE report (softcopies) with all required attachments and cleaned dataset Save the Children Quantitative Performance Evaluation (QPE) SOW 82 4 COMPOSITION OF FINAL QUANTITATIVE PERFORMANCE EVALUATION TEAM The review will be carried out by an external review team6 . The Senior Manager-M&E of Nobo Jibon with the help of the program Chief of Party will coordinate the review activities, with technical assistance from relevant SC/US and SCI units (e.g. food security, livelihoods, M&E). Save the Children in Bangladesh’s MEAL Director will also assist in guiding the review from technical aspect. Position Required Qualifications Responsibilities Team Leader (1) : International Consultant The Team Leader (Masters or PhD) will be contracted for this activity as Team Leader cum evaluation expert. This person should have previous experience in monitoring and evaluation, especially for Title II multi-year food security programming, be fluent in English and have a solid understanding of the relationship between small-scale agricultural production systems, pro-poor market approach, income generation, and behavior change in nutrition/health and household food security. In addition, the team leader must have skills in conducting final survey studies for development programs especially with respect to local institutional capacity building and micro credit, sustainable agricultural development, nutrition, analysis of vulnerability/risk and disaster management. She/he must have practical experience in the areas of designing input masks and data collection tools, developing sampling methods, processing and analysis of data. She/he should master software for processing and analysis of data, specifically SPSS, Epi-Info, Access, etc. Person who could be available in short notice will be an priority for Nobo Jibon. The international consultant will be contracted for this activity as Team Leader Evaluation expert. S/he will have overall responsibility to review documentation, design the fieldwork phase of this exercise and complete the final write up of the report. S/he will also be responsible for field interviews and data validation (if required). S/he will be responsible for sampling, questionnaire coding, database design, training of enumerators, data cleaning, and statistical analysis. The enumerators’ training will be designed by the consultant, at least 5 days will be required, and be implemented in 3 phases: theoretical training; one day of pre-testing; and one day de￾briefing and review of tools. S/he must continue till assignment ends and can’t propose any alternative person in lieu of. Disaster Risk Reduction Consultant (1): (Local or International) • Having regional (Asian) experience working with populations that face regular, rapid onset and chronic disasters. This person should have background in evaluating institutional capacity. LOE is required during questionnaire design, data analysis and reporting. • Person who could be available in short notice will be priority for Nobo Jibon. Reporting to the team leader, s/he will review project documents prior to the fieldwork, participate in the design of interview guides, field interviews regarding DRR. S/he will work in close collaboration with the external consultant on project indicators and key questions related to DRR. S/he must continue till assignment ends and can’t propose any alternative person in lieu of. 6 A short-term contracted consultancy organization. 5 QUANTITATIVE PERFORMANCE EVALUATION MANAGEMENT The QPE team is expected to develop a unique SOW for each of the team members. These SOWs should utilize the questions listed in the following sections as a guide in developing the SOWs for each team member. The Evaluation questions stated at section 2.5 serve as a sample only and should be augmented, based on the expertise of the Consulting Team. 5.1 Team members’ responsibilities and obligations The members shall be responsible for the following: • Adhering to all terms/conditions stipulated in their contracts • Hire local survey firm for conducting quantitative performance evaluation survey through arranging survey equipment (PDA/Smartphone/Tab, Anthropometric measurement tools etc.) • Arrange orientation for the enumerators and facilitate the technical sessions with guidance from Nobo Jibon team (M&E & program). • Determination and documentation of appropriate informed consent practices for survey respondents and clear protocols for voluntary non-participation in line with human subjects protection standards. • Monitor the data collection and ensuring data quality assessment as per USAID & SCI requirement through necessary supervisory hierarchy and technical backstopping. • Assuring the validity of their passports and other relevant travel documentation within Bangladesh • Obtaining their health insurance • Conduct themselves in a respectful manner, while undertaking assignment, which includes not making any promises or commitments to communities and any other persons, on behalf of Save the Children. • Adhering to the agreed time-frames with regard to all activities outlined in the timeline • Providing their own laptop to be used during the assignment (printing, photocopying and other related supports will be provided by Save the Children). 5.2 Nobo Jibon Responsibilities Nobo Jibon program management will be responsible for the following: • Ensure effective coordination of the QPE logistics to facilitate the consultants in undertaking their assignment including their travel, lodging, per-diem and visas. SCI/NJ will not bear any cost related to survey design, implementation, monitoring and reporting. • Provide consultants with all key program documents, including the original proposal and cooperative agreement, the baseline study, midterm evaluation, and other relevant program literature, documentation and reports. • Assist the consultants in scheduling meetings with partners including USAID, Government of Bangladesh line ministries, and other relevant stakeholders • Communicate with the team members Save the Children policies and protocols. 5.3 Ownership of Research Data/Findings All data collected for this review shall remain the property of Save the Children. Any work product resulting from this review must cite the participating partners and USAID as well as include relevant Nobo Jibon staff as a primary or contributing author. 5.4 Budget Budget for the whole consultancy service is mainly divided into three parts. First part includes the consultant cost (Daily Allowance, perdiem for field, transportation (overseas & local), overhead (if any). The following matrix (section-5.8) can give an idea for budgeting for the evaluation task. Second part is for enumerators cost which is related to survey schedule and their level of effort to be planned in lump sum amount for each day work. The rest is for logistics for arranging survey equipment, tools, printing, photocopy, DVDs, internet connectivity, communication and orientation cost for the enumerators and facilitators. A separate cost proposal should be submitted along with technical proposal when bid is open for submission. 5.5 Consultant selection criteria Consultant will be selected based on both Technical and Financial scores and criteria. Evaluation committee members will review the technical proposal first and upon receiving of successful technical proposal, committee will scrutiny the financial proposal of technically eligible firms. The ration of technical and financial proposal will be 70:30 respectively. The following areas will be served as criteria for technical proposal (100 marks) assessment with pass￾mark 60: • Firm’s Previous experience regarding donor, similar work & Bangladesh (20) • Statistically representative Sampling and concurrent with Baseline sampling (20) • Team composition and relevancy (20) • Time-bound rollout plan (10) • Data collection, quality assurance, data management plan & smart tools/sound equipment usage (20) • Analysis plan (10) For financial proposal the following areas should be considered: • Cost Relevancy with technical proposal • Cost-effectiveness • Compliance with SCI procurement policy and payment schedule • Appropriate structure of financial proposal. A financial proposal shall include containing Summary of Costs, Breakdown of Staff Remuneration, Travel and DSAs, miscellaneous, overhead costs (if any) 5.6 Profile of Consultant firm For this consultancy work, the firm should have the following quality: • Any consultancy firm (International) having valid license of doing work with latest tax￾paying evidences. • Team Leader consultant must be international having professional background of more than 15 years in development field. S/he should have access to Bangladesh considering visa issuance. Working experience with USAID is must. • Data collection task should be done by Bangladeshi firm having experience of data collection and management of large project with complex evaluation experiences (more than 10 years). They should have valid registration paper with updated tax-vat Activity Paid Work Days Questions/Comments Team Leader Local Technical Local Consultant/ Rev Fi iew all background documents and all Draft QPE plan t k (protocol)/design, including detailed SoWs for each expert f Design data ent th t ll th ry templates at appropriate hardware (PDA/Smartphone) as per f Ai li d S ti i ppointment of technical resource person & Procure data collection equipment (PDA/Smartphone, Weight scale, height scale, weight, stationeries etc.) Renting cost may be added in addition to the consultants’ LOE if the local consultant firm doesn’t have those equipments tools on their Travel (In country/Overseas) Meeting with team members, Nobo Jibon representatives d USAID registration. Working experience with USAID project would be added quality. For eligibility, Firm and proposed consultants must submit the copy of work order / experience certificate to prove of at least last three (3) international/national experience (whichever applicable). 5.7 Payment process &Schedule The payment will be made as per the Table below upon reaching the milestones. For all payments, Consultant firm has to submit a Letter of Satisfaction and completion of the milestone endorsed by firm and an invoice to SCI BD. Milestone % of Total value Completion of survey design (format, Sampling, work-plan etc.) 20% Completion of the surveys (quantitative & qualitative) 40% Final Report with cleaned datasets 40% Total 100% 5.8 Scheduling with Level of Effort (LOE) The consultancy will commence on approximately 25 Sep, 2014 and proceed until January 15, 2015 as per the timeline outlined at section 3.2: Activity Paid Work Days Questions/Comments Team Leader Local Technical Experts Local Consultant/ Firm Arrange orientation of Enumerators through ensuring field testing the questionnaire, data-entry templates, data entry hardware and measuring equipment etc. Orientation cost with all logistics will be added here with the costing of enumerators LOE for at least 5 days. Conduct data collection and data management for data analysis Enumerators’ data collection time is not estimated here but will be presented separately as lump sum amount. Data analysis & validation Drafting the report and share with all concerned Debriefing meeting with Nobo Jibon, SCI and USAID USAID, NJ, SCI provide feedback on draft report USAID endorses the final review Completing the Final report & shared with all concerned Total estimated Level of Effort (LOE) 87 ANNEX A. Description of Nobo Jibon partnership Save the Children is serving as lead agency overseeing an institutional partnership with well-defined roles, responsibilities and opportunities for sectoral integration and cross-learning. Four local NGOs (GUP, Speed Trust, SAP Bangladesh and CODEC) have been selected as implementing partners (I-Ps). A government agency, the Department of Relief and Rehabilitation, is a partner during construction or rehabilitation of cyclone shelters. Four additional international NGOs – Helen Keller International (HKI), International Development Enterprises (IDE), the WorldFish Center (WFC) and Regional Integrated Multi-Hazard Early Warning System (RIMES) – serve as technical partners (T-P). The table below indicates the partners and roles. Institution Role IMPLEMENTING PARTNERS Community Development Center (CODEC) Multi-sectoral work in Barisal Sadar, Amtali and Taltali Upazilas Gonno Unnuyan Prochesta (GUP) Multi-sectoral work in Mehendiganj and Hizla Upazilas South Asian Partnership (SAP) Bangladesh Multi-sectoral work in Galachipa, Rangabali, Barguna Sadar and Patharghata Upazilas Speed Trust Multi-sectoral work in Dashmina and Kalapara Upazilas Department of Relief and Rehabilitation (DMRD) Collaborate with Save the Children on shelter construction and rehabilitation TECHNICAL PARTNERS Helen Keller International (HKI) Contributions to SO1 (particularly dissemination of Essential Nutrition Action) and gender International Development Enterprises (IDE) Support SO2 market and business development activities World Fish Centre Support SO2 fish production and marketing activities Regional Integrated Multi-Hazard Early Warning System (RIMES) Pilot test advanced cyclone early warning systems. The overall partnership can be characterized as having productive relationships, joint problem solving, and mutual respect. B. Program history and other cross-cutting Nobo Jibon started in June 2010 and its host country agreement was signed in August 2010. The first sub-agreements were signed in October 2010 and the program received monetization commodities in November 2010. A baseline survey was conducted during October-November 2010. Nobo Jibon began work in program communities in December 2010 and distributed first direct distribution commodities to MCHN beneficiaries in February 2011. By the end of 2011, the program had started its work in all upazilas at least with the SO1 interventions. Full coverage was achieved early in 2012. The table below shows the total program coverage to date: Districts Villages VDCs VHCs VDMCs Barguna 441 423 423 410 Barisal 311 292 292 292 Patuakhali 501 441 441 441 Total 1253 1156 1156 1143 Nobo Jibon formed 1156 Village Development Committee (VDCs) and its sub committees like, Village Health Committees (VHC) and Village Disaster Management Committees (VDMC) were formed to represent a cross-section of the community and to give a voice to under-represented groups such as women and extremely poor households. The committees provide a local perspective to decision-making and make valuable contributions to Targeting and Engaging beneficiaries, Phasing in activities, Local Capacity Building, Participatory Monitoring of activities and Sustainability of program interventions. Gender dynamics affect all Nobo Jibon activities and are considered a crucial cross-cutting issue to the program. Issues of gender are important to Nobo Jibon for two main reasons. First, it is well established that women have a vital and essential role to play in maintaining healthy and well-nourished children and families. Second, women in Bangladesh face many disadvantages because of their gender. They are frequently excluded from household decisions, money-making opportunities, and exposure to external messages. Incorporating gender equity into Nobo Jibon is a central goal of the program, cross-cutting all strategic objectives. 13560 women leaders and 4068 male Champions formed and they address gender issues such as intra-household conflict and abuse (in partnership with other group members. Nobo Jibon formed a Gender Working Group and Upazila Gender Focal person who are develop and carry out a comprehensive program-wise gender strategy and implementation. Activities Activities Activities Activities Activities Activities Activities Activities Activities Activities 1.1.1 Develop and 1.2.1 Provide tech 1.3.1 Identify and 2.1.1 2.2.1 Establish 2.3.1 Asset 3.1.1 Facilitate 3.2.1 Construct 3.3.1 3.4.1 Promote A. M&E Plan (will be attached as PDF) B. Nobo Jibon Result Framework Goal: Reduced food insecurity and vulnerability for 191,000 households in nine upazilas of Barisal Division in southern Bangladesh over five years SO1 MCHN Improved health and nutritional status of children U5 and PLW SO2 Market-based Production and Income Generation Poor and extremely poor households have increased production and income SO3 DRR Households in targeted communities protect their lives and assets and quickly resume livelihood activities following natural disasters IR 1.1. – PLW and care-givers of children U5 practice improved MCHN and environ-mental health behaviors IR 1.2. – Households have improved access to integrated health, family planning and nutrition services IR 1.3. –Equity increased within households and communities IR 2.1. – Poor households apply improved knowledge and skills for production and marketing IR 2.2. – Poor households access quality inputs, capital and markets IR 2.3. – Extremely poor households access land, water bodies, and/or productive assets IR 3.1. – Commune-ties manage functional emergency preparedness and response plans IR 3.2. – Commune-ties access appropriate infra-structure for protecting lives and assets in emergencies IR 3.3. – Improved and effective coordination among SC and Nobo Jibon partners to respond to emergencies IR 3.4. – Communities receive and respond to early warning for floods and cyclones implement a BCC strategy 1.1.2 Promote CCM of childhood illness 1.1.3 Promote counseling on health and nutrition. 1.1.4 Provide supp rations to HH with PLW or U2s support to integrate ENA and CCM into public and private health care services 1.2.2 Provide support to health facilities for nut. Info system 1.2.3 Lead advocacy workshops on nutrition-related themes and protocols support women leaders 1.3.2 Form and empower youth groups for community awareness activities 1.3.3 Train male leaders in health, nutrition and social issues Homestead-based experiential learning, demos and input provision 2.1.2 Facilitate cross visits for small producers 2.1.3 Promote selected value chains produce collection points for producers and buyers 2.2.2 Facilitate linkages with input suppliers/ buyers 2.2.3 Facilitate access capital through village savings and loans transfers to targeted HHs 2.3.2 Facilitate access to khas land and water bodies community emergency preparedness and DRR planning 3.1.2 Form and train cadres of community volunteers 3.1.3 Build HH capacity on disaster preparedness 3.1.4 Lead disaster drills/simulations in high risk communities /rehabilitate DRR infrastructure 3.2.2 FFW/ CFW compensation for work on DRR infra-structure Coordinate contingency planning 3.3.2 Build capacity of institutional stakeholders 3.3.3 Train implementing partners in commodity management awareness of GoB early warning system 3.4.2 Pilot test innovations in cyclone early warning 90