1
FINAL EVALUATION 
FOOD FOR EMERGENCY EBOLA VIRUS DISEASE SUPPORT (FEEDS) 
PROJECT IMPLEMENTED IN KAILAHUN DISTRICT BY SAVE THE CHILDREN 
SIERRA LEONE
FINAL REPORT
Save the Children - Sierra Leone
Prepared by: 
MSD Consulting Limited - Sierra Leone
Evaluation Consultants:
Baimba Abdulai Koroma – Team Leader (Food Security and Cash Transfer Evaluation Expert)
Momoh Thomas Bockarie (Socio-economic and Livelihood Expert)
December, 2017
Award Information:
Funded by USAID/Office of Food For Peace; AID-FFP-G-15-00024
2
Acknowledgements
This final evaluation report is made possible by the generous support of the American people 
through the United States Agency for International Development (USAID).
The end-line evaluation team members remain grateful to Save the Children (SC) - Sierra 
Leone program staff and partners for the ownership they exhibited and for providing support 
to MSD through the entire evaluation exercise. To this end, MSD Consulting Limited 
appreciates your inputs into the production of this report. 
Our sincere thanks go to the United States Agency for International Development (USAID) for 
providing the funds to design and implement the FEEDS intervention and its evaluation 
process through SC - Sierra Leone. Without this funding, the FEEDS intervention would have 
not been initiated and subsequently contributed to addressing the acute food insecurity, 
reduced income and malnutrition situations created by the Ebola Virus Disease (EVD) among 
households in Kailahun district. 
We are grateful to the beneficiary participants and implementing partners, including; Ministry 
of Agriculture, Forestry and Food Security (MAFFS), Kailahun district Local Council, National 
Commission for Social Action (NaCSA), Anti Corruption Commission (ACC), representative 
from SPLASH electronic money transfer agency, community chiefs, and Cash Transfer 
Committees (CTC) who participated in this evaluation process by providing relevant 
information used for the production of this report.
We also recognize the important role played by field data collectors whose names are attached 
to this report.
We hope that findings, lessons learned and recommendations generated by conducting this 
end-line evaluation will gainfully be utilized by stakeholders, especially in the design and 
implementation of cash transfer related future program in Sierra Leone and/or other 
developing countries.
DISCLAIMER
The contents of this evaluation report are the responsibility of MSD consulting limited – Sierra 
Leone and do not necessarily reflect the views of USAID or the United States Government.
3
Table of Contents
Acknowledgements 2
Table of Contents 3
Acronyms 4
List of Tables 4
List of Figures 4
Executive Summary 6
Constraining Factors of Project Implementation 9
Lessons Learned 10
1.1 Introduction 12
1.2 Justification of the Final Evaluation 12
2.0 Methodology of the Evaluation 14
2.3 Data Collection 14
2.3.1 Primary Data Collection Tools 14
2.3.2 Secondary Data Collection Tools 15
2.4 Sampling Techniques and Sample Size Determination 16
2.4.1 Sampling Techniques 16
2.4.2 Determination of Sample Size 16
2.5 Data Management and analysis 18
2.5.1 Data capturing and Processing 18
2.5.2 Data Analysis 18
2.6 Data Quality Assurance 19
2.8 Evaluation Ethics 20
2.9 Limitations of the Evaluation 20
3.0 Analysis and Findings 21
3.3.1 Household Dietary Diversity Score (HDDS) 28
3.3.5 Average Household Meal Frequency 34
3.3.6 Utilization of FEEDS Cash by Beneficiaries 34
3.4 Capacity Supports to FEEDS Cash Beneficiaries 36
3.4.1 Capacity Supports to Traders 36
3.4.2 Capacity Supports to Women in Agriculture Groups (WAGs) 38
3.5 Effectiveness of Implementation - Cash Transfer Processes 39
3.7 Intervention Relevance and Strategic Fit 42
3.8 Resource Mobilization and Efficiency of FEEDS Implementation 43
3.9 Partnerships, Collaborations and Coordination 44
3.10 Mainstreaming Gender and Sustainability in FEEDS Project Activities 45
3.11 Monitoring, Evaluation, Accountability and Learning 46
3.12 Constraining and Facilitating Factors of Project Implementation 47
4.0 Lessons Learned and Recommendations 48
4.1 Lessons Learned 48
4.2 Recommendations 49
References 50
Appendices 51
4
Acronyms 
CFSVA Comprehensive Food Security and Vulnerability assessment
CTC Cash Transfer Committees
EVD Ebola Virus Disease
FAO The United Nations Food and Agriculture Organization
FCS Food Consumption Score
FEEDS Food for Emergency Ebola Virus Disease Support
FGDs Focus Group Discussions
HDDS Household Dietary Diversity Score
HHS Household Hunger Scale 
KIIs Key Informants Interviews
MAFFS Ministry of Agriculture, Forestry and Food Security
M&E Monitoring and Evaluation
MDGs Millennium Developments Goals
MSD Management for Sustainable Development
NaCSA National Commission for Social Action
NGOs Non-Governmental Organizations
PDM Post Distribution Monitoring
PPS Probability Proportional to Size 
PRSP Poverty Reduction Strategy Paper
RCSI Reduced Coping Strategies Index 
SDGs Sustainable Development Goals
SLDHS Sierra Leone Demographic Health Survey
SPSS Statistical Package for Social Scientists
UN United Nations
USAID United States Agency for International Development
WASH Water Sanitation and Hygiene
WFP The United Nations World Food Program
WHO World Health Organization
List of Tables
Table 1: Summary of Respondents’ Category and Response Rate
Table 2: Food Consumption Score Benchmarks
Table 3: Meal Frequency (for last 24hrs)
Table 4: Satisfaction with Distance Travelled to Pay Points
Table 5: Adequacy of Beneficiary Cash Amount
Table 6: Satisfaction with Waiting Time to Receive Cash
Table 7: Satisfaction with Safety at pay Points
Table 8: Level of Beneficiary Awareness about the Feedback Mechanisms
List of Figures
Figure 1: Percentage Distribution of Sex of Household Respondents
Figure 2: Percentage Distribution of Age of Household Respondents
Figure 3: Percentage Distribution of Respondents Level of Education
Figure 4: Employment Status of Household Heads
Figure 5: % of Primary Sources of Respondents Household Income 
5
Figure 6: % of Secondary Sources of Respondents Household Income
Figure 7a: % of Tertiary Sources of Respondents Household Income
Figure 7b: % of Overall Sources of Beneficiary Household Income
Figure 8: % of Average Normal Household Expenditure Pattern
Figure 9: Average Household Dietary Diversity Score (HDDS)
Figure 10: % of Respondents Household Dietary Diversity (Multiple Responses)
Figure 11: % of Average Household Hunger Scale
Figure 12: Average Household Food Consumption Score
Figure 13: Average Household Reduced Coping Strategies Index
Figure 14: % of Average FEEDS Cash Utilized on Food and Non-Food Items
Figure 15: % of Food types Purchased with FEEDS Cash (Multiple Responses)
Figure 16: % of Household Member Participation in Decision Making
6
Executive Summary
The May 2014 Ebola Virus Disease (EVD) outbreak in Sierra Leone and the accompanied 
control measures, including quarantine of households, communities, chiefdoms and districts 
restricted the movement of many household members. The restrictions interrupted 
households’ participation in livelihoods and agricultural activities, food production levels and 
access to markets. These impacts eventually worsened households’ economic and food 
security situations, leading to reduced diversified food consumption and increased negative 
coping strategies such as selling household productive assets, child labor, begging, and 
skipping of meals. This left over 88% of Sierra Leonean households food insecure; in Kailahun, 
55.1% of the population was food insecure (9.5% were severely food insecure and 45.6% 
were moderately food insecure). 
The need to complement government efforts to address the acute food insecurity and 
livelihood situation prompted Save the Children, in partnership with its funding agency USAID,
to design and implement a cash-based food program for Emergency Ebola Virus Diseases 
Support (FEEDS) intervention in the nine chiefdoms of Kailahun District. The FEEDS 
intervention was mainly focused on distributing conditional and unconditional cash transfers 
to enable the most vulnerable affected households to recover from the shocks of EVD and 
meet their food security needs.
To measure progress and results achieved, as well as demonstrate learning and 
accountability, Save the Children and USAID commissioned this independent evaluation to 
generate evidence that would permit them conclude what results had been achieved, and to 
enhance future design and implementation of related interventions. The evaluation was 
conducted to:
❖ Assess the extent to which households in targeted district participated in the FEEDS 
intervention to achieve food security as measured by key outcome indicators, including 
household income and livelihoods, household dietary diversity score, household hunger 
scale, household food consumption score, reduced coping strategies index, and 
household meal frequency.
❖ Assess the extent to which conditional and unconditional cash transfers contributed to 
improving the ability of project participants to mitigate, adapt, and recover from the 
economic impacts of the Ebola emergency.
❖ Assess whether new management practices were adopted, and the extent to which 
these affected the activities of female traders and women’s agricultural groups.
❖ Assess the extent to which the systems, processes, and procedures used by FEEDS 
for delivering cash transfers and non-cash assistance to participants were effective, 
efficient, and sustainable during the period of program implementation.
❖ Assess the extent to which collaboration and coordination with public and private sector 
stakeholders contributed towards enhancing effective delivery of cash-based food 
assistance to vulnerable households during the emergency and recovery phases.
❖ Identify lessons learned from the implementation of FEEDS intervention to be used for 
future programming.
A non-experimental pre-test and post-test mixed-method evaluation design with a participatory 
approach was conducted using purposive and three-stage cluster sampling techniques. Both 
quantitative and qualitative data collection methods were used. A mobile-based household 
survey questionnaire designed on the KOBO collect open-source data collection platform was 
used to collect quantitative data on key intervention outcomes and other indicators. Focus 
group discussions (FGD) and key informant interview (KII) guides were used as qualitative 
tools to collect in-depth information, particularly for explanations and triangulation of 
evaluation findings. Quantitative data was analyzed using statistical techniques, which 
7
included tests of significance, while content analysis was used to analyze qualitative data. 
Below are key findings, lessons learned, and recommendations of the evaluation.
Key Findings
Household Income/Livelihood Sources 
❖ Overall, even though casual labor remains the third most prominent source of household 
income, the average reliance of households on crop farming/sales, petty trading, and 
livestock farming as main sources of income have increased. While petty trading 
marginally increased by 0.7%, crop farming increased more substantially, from baseline 
at 32% to an endline value of 58.9%. Livestock farming increased from 3% at baseline to 
6.1% at endline. The evaluation also revealed that casual labor has markedly reduced,
from 22% at baseline to 16% at endline, while other sources of income (begging, firewood 
and charcoal sales, cash for work, or ‘no income’) have largely reduced, from 18.7% to 
6.3% at endline. 
❖ One possible explanation for these findings could be the fact that cash received by 
household beneficiaries has helped households to reduce their reliance on temporary 
casual labor, begging, and remittances. FEEDS cash-supported beneficiaries seemed to 
have focused on engaging themselves in crop and livestock farming on their own plots, as 
well as petty trading activities. They have thus eventually been enhanced to have access 
to better own production and purchase of food from the market in other to meet their food 
security and nutritional needs.
Normal Household Expenditure
❖ Overall, average normal expenditure of beneficiary households on food is seen to have 
increased from 44% at baseline to 45.0% at end-line. Average expenditure on non-food 
items are also noted to have increased. In particular, expenditure on education have 
increased from 10% at baseline to 18.3% at end-line, expenditure on health care increased 
from 8% at baseline to 14.7% at end line, expenditure on agriculture increased from 7%
at baseline to 12% at end-line, and expenditure on savings at had an end-line value of 
4.7% compared to 0% at baseline. Additionally, the prominence of expenditure on other 
items (including household items, rent/shelter, gifts, etc) have been reduced from 15% at 
baseline to 1% at endline, which provides a possible explanation that households now 
spend such moneys on agriculture, food, or savings to improve on their dietary and food 
intake. The increases in expenditure on the above non-food items could be attributed to 
the fact that at baseline, while movement was restricted during Ebola, households were 
found to be recovering from reduced access to their farmlands. Once the restriction was 
over, participants’ household extensive engagement in crop farming for own production of 
food has once again increased. Thus, while diverting excess funds available to 
expenditure on non-food items, beneficiary households still continue to have a 1% 
increase in food expenses, which seem to be used as part of inputs for a more diversified 
food production.
Household Dietary Diversity Score (HDDS)
❖ Overall, the aggregate HDDS determined was 6.3. This value is noted to be greater than 
the FEEDS intervention baseline and target HDDS values of 4.5 and 5.0 respectively, 
which indicates very good achievement as measured from baseline. This is confirmed by 
results from a one-tailed Independent Sample T-test of statistical significance conducted 
at 0.05 significance level and 95% confidence level on average household dietary diversity 
score, which revealed a statistically significant (p<0.001) difference between baseline and 
endline values with respect to average HDDS.
❖ In addition, Chi-Square tests conducted at 0.05 significance level and 95% confidence 
level on each of the individual food groups revealed evidence of a statistically significant 
8
(p≤0.001) difference between baseline and endline values with respect to the individual 
food groups, except for condiments and fat/oil, for which results show existence of 
statistically insignificant differences between baseline and endline for the two groups. 
Household Hunger Scale (HHS)
❖ The HHS scores are categorized into three groups: Little or no hunger, Moderate hunger 
and Severe Hunger. 
❖ Severe hunger among all households has reduced from 1.0% at baseline to 0.5% at 
endline. 
❖ On a whole, moderate hunger in all sampled households markedly reduced from 68% at 
baseline to 34.3% at endline, while little to no hunger largely increased from 32% at 
baseline to 65.2% at endline. This is a huge achievement of the FEEDS intervention. The 
marked reduction in the proportion of beneficiary participants in severe and moderate 
hunger, which was the basis of their selection at baseline, accompanied by the noticeable 
increase in the proportion of households with little to no hunger, could be attributed to 
contributions by the FEEDS intervention, which in effect demonstrates value for money 
with regards the intervention. 
❖ The above findings are supported by Chi-Square tests conducted at 0.05 significance level
and 95% confidence level on three HHS parameters, which revealed the difference 
between baseline and endline values with respect to the three disaggregated categories 
to be statistically significant (p<0.001).
Household Food Consumption Score (FCS)
❖ Measurement of this indicator is based on International Food Consumption Score 
Benchmarks presented in the United Nations World Food Program (WFP) Vulnerability 
Analysis and Mapping Technical guideline (2008), and the Sierra Leone WFP 
Comprehensive Food Security and Vulnerability Analysis (CFSVA) report (2015).
❖ Overall, the average household food consumption score was found to be 52.7 at endline, 
which is considered to be greater than the minimum acceptable consumption of 35.
Household Reduced Coping Strategies Index (RCSI)
❖ The overall average household coping strategies index was found to be 17.5. This is far 
below the average FEEDS intervention target and baseline household coping strategies 
indices of 23.0 and 25.0 respectively. 
❖ The above findings show marked achievement to which the FEEDS intervention would 
have contributed in reducing negative coping strategies among beneficiary households.
❖ The findings are supported by results from an Independent Sample T-test of statistical 
significance conducted at 0.05 significance level and 95% confidence level on an average 
of the five categories/parameters measurement of reduced coping strategies index (rCSI) 
presented in the body of this report. The test revealed a statistically significant (p<0.001) 
difference between baseline and endline values with respect to the average rCSI. This is 
because the 1-tailed significance value (p<0.001) is far less than the significance level 
(0.05). 
Household Meal Frequency
❖ On average, every household member across phases and age brackets was found to have 
eaten one meal or more within the last 24 hours before their participation in the current 
endline survey. 
❖ In the absence of baseline values for comparison, the evaluation team considers this value 
a good achievement. This is evident from taking into account criteria used select target 
beneficiaries and the level of coping strategies beneficiary households employed before 
their selection into the FEEDS intervention. 
❖ Overcoming the difficulties households would have encountered to get a meal per day 
provides evidence that the FEEDS intervention directly improved the frequency of meal 
intake by all categories of household members.
9
FEEDS Cash Utilization by Beneficiaries
❖ The evaluation found that the FEEDS cash was spent on diverse goods and services, 
including food. It was found that out of the total cash given to beneficiaries, 44.3% was 
spent on food while the remaining was spent on several non-food items including 
education (18.0%), health care (15.0%), agriculture (12.0%), and savings (5.7%). 
❖ Regarding the different types of food commodities beneficiaries buy, it was found that 
overall, a minimum of 90% of households spend the cash on rice, fish, vegetable oil/palm 
oil, and condiments. Foods like vegetables, meat, fruits, and eggs did not command much 
expenditure from the FEEDS cash. 
Support to Female Traders and Women in Agriculture Groups (WAGs)
❖ The evaluation found that support given to Female Traders has enhanced their capabilities 
to continue in business and some have even expanded their businesses. Almost all of 
them reported that they have increased their commodity stocking and restocking 
capacities. As a result, most of the WAG-target communities now have almost all the basic 
food stuffs they would need on daily basis. 
❖ The cash and capacity building training support given to the WAGs enhanced their 
capacities to expand cultivation through use of high quality seeds and increased 
knowledge. Their groups have now been strengthened through group management and 
are well positioned to perform better in the coming years. 
Cash Transfer Processes 
❖ The evaluation assessed four key process parameters (distance travelled to pay points, 
cash adequacy, waiting time, and safety at pay points) to help understand beneficiaries’ 
overall satisfaction level about the adequacy of the cash transferred to them and the 
processes that led to the cash distribution itself. Tables 4 to 7 present findings relating to 
this.
❖ It was found that almost all beneficiaries were satisfied with all parameters assessed, 
except cash adequacy, where 62.3% reported that the quarterly cash transferred to them 
was not enough to cover all their food needs for that duration. Reasons given for this 
included large family sizes and the fast rise of food prices during the intervention period, 
which eventually eroded the value of the cash transferred to them at the time. However, 
desk review findings suggest that there was some misunderstanding on the part of the 
beneficiaries, who thought the cash amount given to them was to cover all their food needs 
for a three-month period. In actuality, the cash was supposed to cover only 40% of their 
food needs for that period. This variance in understanding was probably the major reason 
for the level of dissatisfaction reported about the adequacy of the cash transferred. 
❖ Regarding distances travelled to access pay point, 88.8% of respondents reported to be 
satisfied, because the majority of them walked two miles or less to access their designated 
pay points. Only 3.8% reported walking over three miles to their pay point. For waiting 
time, 95.7% reported that they were satisfied with the time taken at pay points before they 
received cash. Most beneficiaries waited for less than an hour to receive cash. Only 5.1% 
reported that they sometimes waited for over two hours before they received cash. For
safety, both at pay points and after leaving with the cash, the evaluation found that almost 
all beneficiaries were satisfied with this, at 97.4% and 95.1% respectively. Even when they 
got back to their communities, the evaluation assessed whether beneficiaries received any 
pressure from community leaders or CTC members to give part of the cash to them. 99.4% 
reported never experiencing that.
Constraining Factors of Project Implementation
❖ Bad road networks and poor mobile communications networks slowed down the pace of 
activities, including the cash transfer itself as well as monitoring activities.
❖ Some Cash Transfer Committee (CTC) members reported that in some cases they felt the 
urge to quit the project because ‘they were not getting any direct benefit (tokens, etc.) in 
10
return for their services’. This was confirmed from KIIs with community leaders and field 
program staff, who reported that they sometimes had to spend a long time encouraging 
CTC members to continue working and not quit. According to them, it would have taken 
too much time to get new CTC members onboard and get them understand the 
intervention. Rather, they were encouraged to maintain current CTC members partly 
because the CTC members involved already understood the project concept and partly 
because they already knew the cash beneficiaries. This problem of CTC members quitting 
the project was largely mitigated by giving them FEEDS customized T-shirts. 
Facilitating Factors of Project Implementation
❖ The FEEDS project had enough staff to execute all key functions. Assigning specific staff 
to specific chiefdoms ensured that the project team was always close to the beneficiaries. 
This enhanced monitoring activities and flagged up issues that were addressed promptly.
❖ The evaluation noted that all the partners that the FEEDS team worked with (including the 
CTC and local authorities) performed their respective duties. This enhanced the level of 
results that the evaluation found. 
Note: Though some members of the CTC sometimes complained of “no tokens”, their 
involvement played a very crucial role in achieving the overall result that the FEEDS project 
produced.
❖ The project was well funded. This was a huge source of timely and full implementation of 
activities, which permitted the achievement of desired results.
❖ Effective sensitizations and house-to-house visits by field staff on the correct use of cash 
received, coupled with the complementary trainings, ensured that beneficiaries used more 
of the cash to access food items, rather than non-food items. This helped improve their 
food security statuses, as evidenced by changes in the respective indicators. The special 
attention paid to household decision-making regarding the use of FEEDS cash also helped 
avert most potential conflicts, which may have been the source of the level of women 
empowerment the evaluation found. 
Lessons Learned
❖ When the income of households increases in cash, they are motivated to increase their 
expenditure on other important items that were not previously prioritized as expenditure 
items when there was less or no income in the household. The FEEDS cash expenditure 
patterns seem to be contrary to the expectations of the FEEDS intervention that at the 
end of the intervention, 70% of the EFSP cash received by households would be spent 
on food items. However, the observed decrease in EFSP cash expenditure on food 
(44.3%) was compensated for by expenditures on other important and productive 
household needs like education, health care, agriculture, and savings, all of which 
contribute to long-term household food security. 
❖ Save the Children deployed staff directly in the target chiefdoms. This provided for a 
closer link between implementing staff and the beneficiaries throughout the project. This 
was very helpful for effectively monitoring market prices, cash utilization, and nutrition￾sensitive messaging. This strategy created high SC/USAID visibility, which in turn helped 
improve implementation through prompt and firsthand information gathering that was 
used to make informed decisions about the project.
❖ Establishing the right partnerships with state and non-state actors adds special value to 
quick-impact projects like cash transfer projects. The presence of NaCSA, ACC, KDC,
and CTCs added integrity and visibility to the whole process. 
❖ When the right beneficiary targeting is done, cash interventions could have a very big 
impact on the daily lives of the beneficiaries. Their food security situation will be 
significantly impacted in the shortest possible time, as measured by the statistically 
significant changes that occurred in the values of target household’s HDDS, HHS, and 
rCSI scores.
11
❖ The involvement of CTCs created a meaningful impact, especially in targeting the right 
beneficiaries and conveying information quickly to beneficiary households. However, 
when they are not allowed to benefit directly, they will have problems with total 
commitment to the program.
❖ In the absence of effective mobile technology, a well-deployed and monitored offline cash 
transfer project can adequately replace technology-based cash transfers. 
Recommendations 
❖ The average meal frequency of households per day was noted to be approximately two 
during the final evaluation. There is need for this figure to reach a minimum of three for 
better nutrition outcomes. Therefore, the Government of Sierra Leone through the 
Ministry of Health and Sanitation, represented by the District Health Management Team 
(DHMT) and other related community structures, should continue with activity-based and 
monitoring program strategies to increase or maintain the community’s access to 
sufficient nutritious foods in the right varieties.
❖ The evaluation found that many households have diversified their diets because of the 
EFSP intervention. This has prompted most to start or increase their backyard gardening 
to have continuous access to fresh vegetables, fruits etc. On that note, the Government 
of Sierra Leone, through the Ministry of Agriculture, Forestry and Food Security (MAFFS),
should undertake regular sensitizations on the benefits of backyard gardening and the 
importance of eating these foods to help improve household nutritional status. 
Additionally, MAFFS block extension officers should continue monitoring the participants 
who were trained in post-harvest management technology to ensure they continue to 
employ it in their farming activities. They should also provide technical assistance where 
necessary. 
❖ The cash transfer processes and systems were largely effective, as noted by the 
beneficiary’s satisfaction levels with the distances covered to access pay points, waiting 
time to receive cash, and safety both at pay points and after leaving pay points. It is 
recommended that the strategies employed to achieve these must be maintained by 
Save the Children in similar future interventions.
❖ The direct deployment of Save the Children staff in the target chiefdom helped 
significantly in achieving results. This strategy should be replicated and/or scaled up in 
future similar interventions.
❖ There was some level of misunderstanding on the part of the beneficiaries about the 
percentage of household food needs the transferred cash was supposed to cover. Most
thought the cash was meant to cover all the household food needs. Therefore, most 
households (62.1%) reported that the cash was not adequate for their household food 
needs. It is recommended that in the future, a very clear sensitization must precede the 
transfer of cash, which should continue throughout the life of the project.
❖ The evaluation noted that there was limited involvement of the private sector in the overall 
implementation. Though there was not observed direct negative impact on the project, 
the high potential of this sector could be tapped into if more organizations from the sector 
were drawn into the project. We therefore recommend that in similar future projects, Save 
the Children should partner with more than one private sector organization. 
12
1.1 Introduction
The May 2014 unprecedented Ebola Virus Disease (EVD) outbreak in Sierra Leone resulted 
in a worsening of the economic and food security situations for households in the country. To 
contain the rapid spread of the disease, control measures were identified and implemented by 
government and its development partners. These control measures, including quarantine of 
households, communities, chiefdoms, and districts graduated into restricting movement of 
household members from one point to the other. These measures interrupted household 
access to markets and negatively impacted their investment and participation in groups and 
agricultural activities. Consequently, households suffered reduced household economic 
access to diversified food and to food production level. This eventually resulted in reduced 
food consumption stability of households over time, increased negative coping strategies such 
as selling of household productive assets, child labor, begging, and skipping meals. These 
dismal situations left Sierra Leone with only about 11.2% of all poor households’ food secure, 
and only 3.3% in Kailahun district (CFSVA, 2015). 
The need to urgently address the acute food insecurity situation prompted the identification, 
designing, and implementation of cash transfer intervention, which was intended to contribute 
to improved food security situations among vulnerable beneficiaries. Save the Children, in 
partnership with USAID, took the challenge to complement government effort towards 
achieving implementation of the National Ebola Recovery Strategy through cash-based Food 
for Emergency Ebola Virus Diseases Support (FEEDS) intervention in nine chiefdoms in 
Kailahun District, (Dea, Njaluahun, Jawei, Kpeje Bongre, Kissi Teng, Kissi Kama, Luawa, 
Malema and Upper Bambara). 
The FEEDS project, implemented in three different phases (reaching a total of 77,211 
individuals in FY 2017 and an estimated 91,154 over the life of the award) was mainly focused 
on distributing conditional and unconditional cash transfers to enable the most vulnerable 
affected EVD households recovered from the shocks of Ebola and meet their food and nutrition 
needs. As implementation of the project came to an end, SC commissioned an independent 
evaluation to assess the impact and extent to which results were achieved, contracting MSD 
Consulting Limited to undertake the current final evaluation to which this report serves as a 
primary deliverable.
This report is a key deliverable for the endline evaluation exercise. After the introduction, 
sections on the evaluation objectives and justification, methodology, analysis of findings, and 
lessons learned and recommendations are presented.
1.2 Justification of the Final Evaluation
In line with standards and best practices emerging from the 2005 Paris Declaration and 2008 
Accra Agenda for Action on enhancing development effectiveness, the cash-based FEEDS 
final evaluation demonstrates learning and accountability to stakeholders for SC services 
delivered to beneficiaries. This permits project stakeholders to make conclusions on 
achievement (or non-achievement) of intervention results based on evidence. The evaluation 
findings allow SC, its funding partner, and other stakeholders to properly plan, implement, 
monitor, and evaluate future related food security-based cash transfer interventions to address 
issues affecting vulnerable/marginalized groups in emergencies.
1.3 Purpose and Objectives of the End-line Evaluation
The overall objective of the end-line evaluation was to conduct a performance evaluation of 
the Save the Children USAID-funded FEEDS program in other to assess its effectiveness, 
efficiency, sustainability, and impact on improving the food security of beneficiary households 
13
in an emergency context after the EVD outbreak. The endline quantitative survey primarily 
focused on assessing the following key indicators for a sample of unconditional cash transfer 
beneficiary households: (1) income and livelihood status, (2) the household dietary diversity 
score, (3) the household hunger scale, (4) the household food consumption score, (5) the 
reduced coping strategies index, and (6) the household meal frequency. The end-line 
evaluation also assessed whether new management practices were adopted and the extent 
to which these have affected the sales of two groups of beneficiaries, namely female traders 
and the women’s agricultural groups. Information gathered from the endline evaluation was 
used to measure the FEEDS program achievements over the course of its implementation 
period and report the results against predefined indicators set forth in the agreement between 
USAID and Save the Children. Additionally, the final evaluation provides documented learning 
evidence on SC implementation approaches and will inform future programming of similar 
interventions.
The specific objectives of the endline evaluation were to answer the following questions:
1) To what extent did households that participated in the FEEDS program achieve food 
security and improved resilience as measured by the following quantitative variables?
a. Income and Livelihood
b. The household hunger scale
c. Household dietary diversity score
d. Reduced coping strategies index
e. Food consumption score
f. Meal frequency
In addition to analysis of the quantitative data, the evaluator was expected to address the 
extent to which the below components contributed to improving the ability of project 
participants to mitigate, adapt, and recover from the economic impacts of the Ebola 
emergency:
● Unconditional cash transfers
● Conditional cash transfer support to Female Traders
● Conditional cash transfer support to Women in Agriculture
● Agronomic and financial management / business skill training 
2) To what extent were the systems, processes, and procedures used by FEEDS for 
delivering cash transfers and non-cash assistance to participants effective, efficient, and 
sustainable during the period of program implementation? In answering this question, the 
evaluator should address the following:
a) Accountability, timeliness, and cost of delivering the cash and non-cash assistance to 
participating households
b) Assess the participants’ perception about the cash transfer and non-cash delivery 
processes and procedures
c) Explore the gender implications of cash transfer and non-cash assistance relating to
o Decision making
o Gender roles and responsibilities
o Gender based violence 
d) What were the key lessons learnt during the FEEDS program implementation?
3) To what extent did the collaboration and coordination with public and private sector 
stakeholders resulted in enhanced capacity to deliver cash transfers as a form of social 
14
protection and food assistance to vulnerable households during emergency and recovery 
phases? In answering this question, the evaluator should address the following:
a. Connectedness and coherence with other actors such as other private sector, 
NGOs, Donors, UN organizations, and multilateral agencies.
2.0 Methodology of the Evaluation 
2.1 Approach
A performance and summative evaluation with mixed and participatory approaches were used 
in this endline evaluation. The approaches permit stakeholder involvement as well as the use 
of quantitative and qualitative data collection techniques and tools. In addition, these 
approaches ensured effective triangulation of information and generalization of the FEEDS 
evaluation findings.
2.2 Evaluation Design 
A non-experimental pre-test and post-test mixed-method evaluation design with a participatory 
approach was used. The design allowed effective triangulation of information gathered from 
different stakeholders. Since project implementation was done in phases, the evaluation 
design strategies used accordingly took into account variations in the different extension 
phases of the EFSP intervention. This design provides an approximate estimate of intervention 
impacts in terms of changes in outcome indicators over the project lifecycle. If carefully 
implemented, significant difference will be obtained in each test; and this consequently would 
provide evidence of the intervention effects.
2.3 Data Collection
The designed FEEDS logical framework, progress monitoring report and other documents, 
evaluation objectives, and data collection tools used at phase I baseline and midline were 
used to guide the current evaluation exercise. These were used to develop data collection 
tools to assess the extent to which project objectives were achieved, and consequently to
determine the performance of the FEEDS intervention in Kailahun district. The content of 
phase I baseline and midline tools were largely maintained, especially for the key outcome 
indicators, to allow comparison of baseline, midline and endline values.
A combination of quantitative and qualitative methods was employed to collect data using the 
different data collection tools from the respective sources. Both primary and secondary data 
were collected. Secondary data was collected by document review. Primary data was 
collected using a quantitative online mobile-based household survey questionnaire designed 
on KOBO software platform and qualitative tools, including focus group discussions and key 
informant interview guides. Beneficiary (household) surveys were used by enumerators to 
collect data from randomly selected beneficiaries in randomly selected clusters and respective 
communities. Qualitative data was used to collect data from randomly selected cluster 
communities across targeted chiefdoms in the FEEDS intervention. Descriptions of both 
primary and secondary data collection tools used are presented below:
2.3.1 Primary Data Collection Tools 
❖ Online Mobile-based Household Survey Questionnaire:
The questionnaire was designed on the Kobo software platform and administered to direct 
project beneficiaries. The tool was used to collect data on the following indicators, 
including; household demographic characteristics, household income and livelihoods, 
expenditure patterns (on food and non-food items), changes in agricultural and financial 
15
assets owned by households, household hunger scale, household dietary diversity score, 
food consumption score, reduced coping strategies index, household meal frequency, 
cash transfer processes and utilization, capacity building and practices (on agronomic 
practices, nutrition, hygiene behavior, and financial/business management skills), 
relevance of FEEDS intervention to beneficiaries and communities, monitoring and 
accountability, and gender integration into the FEEDS intervention planning and 
implementation and its related effects. On average, administering the tool to one 
beneficiary respondent took 25 minutes. 
❖ Focus Group Discussion (FGD): 
This was a group discussion guide designed and used to collect detailed information from 
direct SC FEEDS beneficiaries in small groups of 8 – 10 participants each. The guide was 
used to conduct group discussions separately for men and women. These included direct 
unconditional cash transfer beneficiaries, women farmers (agricultural), and small-scale 
traders who benefitted from the FEEDS intervention. The tool was comprised of open￾ended questions, administered by trained enumerators, to have respondents freely 
express their opinions and perceptions about the FEEDS intervention. The tool collected 
data on household income and livelihoods, household expenditure (on food and non-food 
items), relevance of FEEDS intervention, capacity building trainings and practices (on 
agronomic practices, nutrition, hygiene related issues, and financial/business 
management skills), cash transfer processes and utilization, monitoring and accountability, 
and coordination and sustainability of the FEEDS intervention.
❖ Key Informant Interview (KII) Guide: 
This guide was designed and used to collect detailed information from different categories 
of individual respondents, including community chiefs, Local Council, Ministry of 
Agriculture, Forestry and Food Security (MAFFS) at district level, National Commission for 
Social Action (NaCSA), Anti Corruption Commission (ACC), representatives from SPLASH 
electronic money transfer agency, Community Identification Committees (CICs) or cash 
transfer committee members, SC FEEDS intervention staff, and other partners. The tools 
comprised of open-ended questions, administered to the above stakeholders by trained 
data collectors to collect detailed information on relevance of FEEDS intervention (to 
beneficiaries, communities, and country strategic priorities), efficiency of the intervention, 
cash transfer processes and utilization, monitoring and accountability, impacts of the 
intervention, coordination and sustainability of the intervention, and successes, 
challenges, and lessons learnt. 
2.3.2 Secondary Data Collection Tools 
The major tool used here was desk review of documents, mainly to collect secondary data 
from FEEDS intervention documents and other existing studies and literature relevant to 
conditional and unconditional cash transfers, as well as food security (particularly in 
emergencies) across developing countries including Sierra Leone. 
FEEDS collected and reported data periodically on the following indicators, although the 
definition and methodology of collection and analysis varied slightly for an emergency 
context: 
- FFP#28 Prevalence of households with moderate or severe hunger (Household 
Hunger Scale)
- FFP#29 Average Household Dietary Diversity Score (HDDS)
- FFP#16 Value of incremental sales attributed to program implementation
- FFP#11 Number of individuals who have received USG supported short‐term 
agricultural sector productivity or food security training
16
-FFP#9 Number of farmers and others who have applied new technologies or 
management practices as a result of USG assistance.
Documents prioritized for review included: SC FEEDS narrative proposal and Save the 
Children country strategic plan, FEEDS needs assessment report, baseline(s) and mid-line 
evaluation reports, project internal routine and annual monitoring reports, the Sierra Leone 
Smallholder Commercialization Program, Post Ebola Recovery Plan, Comprehensive Food 
Security Vulnerability Assessment (CSFVA, 2015), cash based FEEDS implementation and 
impact documents (including best practices and lessons learnt), post distribution and market 
monitoring reports, PRSP II & III, emergency food security technical and monitoring 
documents by USAID, World Food Program (WFP) and FAO and other organizational 
documents on household dietary diversity score, food consumption score, household hunger 
scale, meal frequency, and reduced coping strategies index. 
2.4 Sampling Techniques and Sample Size Determination
2.4.1 Sampling Techniques
At baseline (phase I) and midline (phases I and II), a stratified random sampling was followed, 
using a statistical formula developed by Krejcie and Morgan (1970) to determine 
representative sample sizes. This sampling was not, however, based on a phased-program 
design. Per the CDC guidance for comparable evaluations, the FEEDS program stratified each 
sample for the Phase I baseline survey, midline of Phase I and II survey, and baseline of 
Phase III survey per chiefdom for all three datasets considered. Since the phased-program 
design and implementation within the life cycle of the FEEDS intervention only emerged after 
the phase I baseline and the midline were conducted, conducting a pipeline design final 
evaluation would be challenging. Additionally, using the Krejcie and Morgan (1970) formula 
for each phase’s endline seemed to be impractical given the time and resource limitations for 
the final evaluation.
2.4.2 Determination of Sample Size 
Given the sampling challenges described above, the evaluation team used a statistical formula 
developed by Magnani (1999) presented as addendum in FANTA III Sampling Guide (2012) 
at 95% confidence level to determine a representative survey sample size of household 
beneficiary participants to whom the online mobile-based quantitative survey data collection 
tool was administered. The sample size determination took into account the design effect and 
key intervention outcomes indicators for which adequate information was available. Below is 
the sample size formula used:
Where,
n = required sample size of cash beneficiary respondents
D = Design effect (for a two stage cluster sampling, D=2.0 for FFP Program)
P1 = the value of the key indicator at baseline (or a proxy value), expressed as a 
proportion between 0 and 1
17
P2 = the planned target value of the key indicator at the end-line/final evaluation, 
expressed as a proportion between 0 and 1
Zα = the Z-score corresponding to the desired confidence level (α = 1 - a). At 95% 
confidence level, a = 0.05, which gives α =0.95, and Zα = 1.645
Zβ = the Z-score corresponding to the desired statistical power (β = 1 - b). Consistent 
with FANTA III Sampling Guide (2012), for FFP program, b is typically set at 
b = 0.2, which gives β=0.80 and Zβ = 0.840
❖ Design effect is the factor by which the sample size for a cluster sample would have to 
be increased to produce survey estimates with the same precision as a simple random 
sample.
Based on information provided about baseline and target values for the three key indicators 
[Household Hunger Scale (HHS), Household Dietary Diversity Score (HDDS), and reduced 
Coping Strategies Index (rCSI)], and keeping a large enough number of clusters to provide 
representative sample size, a mix of sample sizes was determined. For HHS, very small non￾representative sample size was determined (46 sample respondents), while very large sample 
sizes (almost census values) were determined for both HDDS and rCSI (3,443 and 4,932
sample respondents respectively). With these determined sample sizes, in addition to the 
limited time and other resources, the evaluation team assumed at least a 10% estimated 
change in each of the key outcome indicators during the life span of the project. This allowed 
for the determination of a minimum required sample size of 605 beneficiary participants. 
However, accounting for non-response participants and sample representation, provisions 
were made for an additional 48 response participants for flexibility.
In line with the principle of Probability Proportional to Size (PPS) and consistent with the 
number of households served in the three FEEDS phases with three different groups of 
beneficiaries across chiefdoms and respective cluster points, the above sample size was 
proportionally distributed among these phased groups as follows:
Phase I: 6,445 Beneficiary unique households
Phase II: 1,492 Beneficiary unique households
Phase III: 7,194 Beneficiary unique households1
According to Magnani (1999) as Addendum in FANTA III Sampling Guide (2012), and as used 
in this evaluation, a minimum of 30 clusters represents a figure adequate to ensure that 
sample target groups are sufficiently spread across enough clusters of intervention coverage 
area. In this evaluation, a total of 33 randomly selected clusters out of the 77 clusters for the 
FEEDS intervention with a minimum of 21 beneficiary respondents per cluster was used. In 
addition, three (3) communities per cluster with a minimum of seven (7) household beneficiary 
participants (per community) were randomly selected and surveyed. 
Also, a total of 22 selected focus group discussions (one per randomly selected cluster) of 8 
to 10 participants per group were selected and conducted across the intervention chiefdoms. 
These included six (6) FGDs each for women and men beneficiaries participants, five (5) for 
traders, and five (5) for women in agriculture were conducted across the intervention coverage 
 
1 The additional households, that entered the program in Phase I and II, but who were also benefitting 
in Phase III, were not in the list of unique Phase III households. 
18
area. FGDs for the different groups were conducted in same randomly selected communities. 
Additionally, 51 key informant interviews were conducted with the different categories of 
individual respondents identified earlier. In particular, KII for Cash Transfer Committees
(CTCs) and community chiefs were conducted in same randomly selected communities where 
the FGDs were conducted. 
Below is a table, presented by tools, a summary of respondent category and response rate.
Table 1: Summary of Respondents’ Category and Response Rate
Respondent Type Expected and Actual Number of Tools Administered 
Questionnaire KIIs FGDs
Expected Actual Expected Actual Expected Actual
Cash Participants 605 653 - - 22 22
Local Council - - 1 1 - -
MAFFS Reps - - 1 1 - -
Community Chiefs - - 22 22 - -
Anti Corruption - - 1 1 - -
Community Transfer 
Committees (CTCs)
- - 22 22
NaCSA - - 1 1 - -
SPLASH - - 1 1 - -
FEEDS Program
Management Staff
- - 2 2 - -
2.5 Data Management and analysis
2.5.1 Data capturing and Processing 
Quantitative data was collected through the online mobile-based Kobo data collection system 
with incorporated skip logic to enhance data accuracy. The data was downloaded through MS 
excel and subsequently imported into SPSS Version 23.0 database, where it was cleaned to 
avoid duplication and other errors to ensure quality. 
Qualitative data was collected and captured using paper and, in some cases, recorders. The 
recorded data was transcribed and merged with paper-recorded data. Data was captured by 
thematic indicators and entered into MS Word by themes, response category, and chiefdom. 
This made it easier for data screening to ensure quality assurance with regards accuracy and 
consistency. 
2.5.2 Data Analysis
Quantitative data was analyzed by gender (sex) using SPSS version 23.0, which allowed for 
descriptive statistical analysis and presentation of evaluation findings. For selected outcome 
indicators (HDDS, rCSI, and HHS) with available baseline datasets, Chi-Square statistical 
tests were performed to estimate, based on changes observed, whether there were 
statistically significant differences between endline and baseline values particularly for HHS. 
19
In addition, a one-tailed Independent Sample T-test was performed for average HDDS and 
rCSI, to estimate, based on changes observed, whether there were statistically significant 
differences between endline and baseline values with respect to the mean value of the 
indicator.
Qualitative data was analyzed using thematic content analysis of evaluation information 
gathered from identified sources, which took into account analysis framework and plans
agreed upon by an evaluation team and Save the Children. The analysis also took into account 
identification of emerging themes and establishment of relationships between these themes. 
The data was analyzed and findings presented within the framework of household 
demographic characteristics, household income and livelihoods, intervention relevance and 
appropriateness of its design, efficiency of resource utilization, and effectiveness and impact 
of the intervention. Effectiveness and impacts were analyzed and discussed with specific 
reference to Household Dietary Diversity Score (HDDS), Household Hunger Scale (HHs), 
reduced Coping Strategies Index (rCSI), FEEDS cash transfer implementation processes, 
cash utilization, and technical assistance provided by FEEDS and the accompanied 
beneficiary practices. In the absence of baseline values, additional food security indicators 
that are key to the FEEDS intervention, including Household Meal Frequency and Food 
Consumption Score (FCS), were also analyzed. The analysis framework took into account 
presentation of endline findings on partnership and collaboration, mainstreaming of gender 
and sustainability into the intervention, and monitoring, evaluation, accountability, and learning 
of the intervention. 
2.6 Data Quality Assurance
For quality and safety of data collected, the evaluation team established reporting lines of 
supervisors and enumerators assigned to them. This maintained a free flow of communication 
between enumerators and supervisors and between them and lead consultants, through which 
barriers encountered were easily overcome. Daily summary report forms were prepared and 
given to supervisors to be completed by enumerators and submitted on a daily basis to 
supervisors before the start of next day data collection. Additionally, provisions were made on 
the Kobo platform for names of enumerators and supervisors to enhance the ability to easily 
trace any anomaly. 
To verify the presence of data collectors in their designated locations, lead consultants were 
constantly engaged in random spot checks on supervisors and enumerators, visiting some of 
the beneficiary households where administration of survey tools was ongoing or had already 
been administered. This was to enhance data quality through monitoring of field data collection 
and to ascertain that data collectors actually visited identified sample respondents as planned. 
In few cases, callbacks were made at households where the FEEDS intervention beneficiaries 
were not met on first visits. However, data collectors were successful in those few cases.
At the end of field exercise, the downloaded data from Kobo into MS Excel was subsequently 
exported into SPSS version 23.0. Consistent with SC data quality protocols, the following 
steps were taken to clean the quantitative data:
➢ Saved the original (downloaded) excel data file and created a copy of the downloaded 
data to serve as reference point.
➢ Checked to verify beneficiary participants’ full name, location, and their Segovia ID 
number given to them for receipt of cash. This was to ensure that data was collected 
from direct beneficiaries.
➢ Added a filter to the top column of the downloaded dataset and used the filter to check 
for beneficiary consent status, particularly for those that did not give consent to 
provide information. This was done to follow up with any data collector who could 
have collected information without the respondent’s consent.
20
➢ A further check was done by running frequencies for beneficiary response rates to 
confirm the sample representativeness, particularly for the calculated minimum 
sample size that was determined to be adequate for the endline evaluation, so that 
inferences drawn from the large enough sample size would be meaningful.
➢ Data was then exported into SPSS version 23.0 for further cleaning, followed by 
analysis.
➢ During data analysis, there were cases where average comprehensive values of 
indicators needed to be calculated. In those cases, the required portion of the data 
set was carefully copied into excel from SPSS and subsequently create columns for 
total, averages and percentages. 
2.7 Recruitment and Training of Data Collectors
Fifteen experienced data collectors (supervisors and enumerators) who have participated in 
several prior evaluation data collections, including working with MSD were recruited. A two￾day training, facilitated by lead consultants and one SC FEEDS staff, was conducted for all 
data collectors, who were trained on the background overview of the FEEDS intervention, 
evaluation design, sampling techniques used, and evaluation ethics. Facilitators took trainees 
through each of the data collection tools, and trainees were divided into smaller groups for 
role-play sessions. This was followed by field pre-testing of the tools. In the role-play sessions 
and field pre-testing exercises, enumerators were assessed on the practical use of the tools 
for effective and ethical data collection from respondents.
2.8 Evaluation Ethics
Development research ethics were maintained throughout the entire evaluation process. In 
this way, written or verbal informed consent from evaluation participants was obtained. FGD 
and individual survey participants were also told the confidentiality of their names and 
information would not be shared with unauthorized users. Furthermore, participants were 
instructed by the evaluation team that they (evaluation participants) were expected to be 
confidential about all that was said during interviews and discussions. In addition, the rights of 
respondents to end their participation in the evaluation at any point during the course of 
information gathering was made clear to them. The evaluation team constituted of both males
and females to take into account asking questions dealing with sensitive topics where 
necessary, from which participants’ total anonymity was assured. Linked with SC ethical rules, 
complaint mechanisms were also established with identified subjects to contact in events 
where any concerns were to be raised. Although no sensitive cases were identified, referral 
form templates were prepared for completion by enumerators on sensitive cases or serious 
child protection issues, to be submitted to supervisors for onward submission to lead 
consultants and eventually to SC. 
2.9 Limitations of the Evaluation
1. The evaluation team used only a 10% estimated change in each of the key outcome 
indicators during the life span of the project, rather than the targets set by the project. 
This provided the team with a large minimum required sample size of intervention 
beneficiaries. The sample size for the evaluation was larger (653) than that used for 
phase I and III baselines and midline (400 each), but was comparatively smaller than 
having an endline with 400 households for each of the intervention phases. This was 
because the calculation using the statistical formula developed by Magnani (1999), and 
using project determined baseline and target information, resulted in sample sizes that 
were on one hand too small and on the other hand too large relative to the number of 
targeted beneficiaries. See Section 2.4.2 for details.
21
2. Chi-Square tests were only conducted for HHS, and t tests for HDDS and rCSI outcome 
indicators, to test whether there are statistically significant differences between baseline 
and endline values with respect to these indicators. No statistical tests were run on 
income and livelihoods data. Furthermore, the baseline datasets did not include Food
Consumption Score and Meal Frequency. This limited the evaluation team to conducting
similar Chi-Square and t tests to estimate and ascertain whether the associated changes 
in values of these indicators from baseline to endline were statistically significant. 
3. The absence of baseline data for certain outcome indicators made it impossible for the 
evaluation team to have perform a direct comparison of baseline and endline indicator 
values. The evaluation could therefore only compare these endline values with national 
survey data, which likely resulted in an under/over estimation of attributable changes 
estimated in this report. 
4. Field workers were challenged with accessing randomly selected target communities due 
to poor road networks and crossing of rivers/streams. In these instances, data collectors 
are strictly advised by MSD Consulting Limited not to cross streams/rivers, so they had 
to travel long distances to access the randomly selected survey communities and 
respective beneficiaries, which took more time and resources.
3.0 Analysis and Findings
3.1 Household Demographic Characteristics
● Sex and Age of Household Beneficiary Respondents 
A total sample of 653 beneficiary respondents were randomly selected and participated in the 
survey. Figure 1, below, shows that among these respondents, more females were surveyed 
(72.4%) than their male counterparts (27.6%). Evaluation respondents are thus dominated by 
females.
Figure 1: Percentage Distribution of Sex of Household Respondents
In Figure 2, it can be seen that over 98% of all household respondents that participated in this 
end-line survey fall in the adult age range of 25 to 50 years and beyond. As revealed in the 
figure below, more respondents within the age range of 25-49 years were selected as FEEDS 
participants. 
22
Figure 2: Percentage Distribution of Age of Household Respondents
● Average Households Size of Beneficiary Participants
Overall, the average household size of survey respondents is 5.7. Comparatively, the value is 
consistent with the Sierra Leone CFSVA (2015) and the 2013 Sierra Leone Demographic 
Household Survey results at both national and district levels, especially Kailahun District 
where this end-line evaluation was conducted. 
● Educational Level of Household Respondents
This is presented in Figure 3 below. 
Figure 3: Percentage Distribution of Respondents Level of Education
It can be seen from the Figure 3 above that the vast majority of household participants have 
no formal education; over 80% of the respondents in this evaluation fell in the category of 
participants with no formal education. Figure 3 clearly shows that next to the high proportion 
of respondents with no formal education are respondents who could only be educated up to 
primary level (8.3% of all respondents) and others who could attain some form of koranic 
23
education (4.9%). Despite the above, a minimal proportion of respondents were found with 
Junior Secondary School levels of education (3.8%). 
Of the total respondents, about 64.9% (424) of females and 15.9% (104) of males do not have 
any formal education. The findings indicate that women remain more vulnerable than men in 
terms of education out of the FEEDS beneficiary participants. The low level of education 
among women is evidence of their increased vulnerability to poverty and limited access to 
resources in these settings.
3.2 Household Socio-economic Characteristics
3.2.1 Income and Livelihood Sources
● Employment Status of Household Heads
Figure 4 presents the employment status of household heads. The overall average shows that 
almost all household heads are either unemployed (57.9%) or self-employed (41.7%). No 
meaningful evidence of employment in formal institutions is observed in this evaluation. 
Although evidence of changes that could be attributed to the intervention with regards to a 
reduction in unemployment and an increase in self-employment of household heads cannot 
directly be determined in this evaluation, connecting these evidences to baseline findings on 
sources of household income may yield interesting results that resonate with the above 
findings. At baseline, most respondents across study chiefdoms were found selling some of 
their crops, as well as engaging in casual labor and petty trading in the informal markets for 
dependable income sources. 
Figure 4: Employment Status of Household Heads
● Household Income Sources 
The evaluation revealed that beneficiary households depend on diverse sources of income. 
Presented below are sections of findings on three main (primary, secondary, and tertiary) 
income sources. Findings presented in this section on proportions of household income used 
data that was analyzed outside of cash transfers, to ensure comparison of endline and 
baseline values. 
Primary Income Sources: Figure 5 presents findings on beneficiary household primary 
sources of income. As revealed in the figure, crop farming/sales remain the major primary 
income source among households in the FEEDS intervention coverage area. 89.5% of 
beneficiaries reported that they rely on crop farming/sales. 
24
Figure 5: % of Primary Sources of Beneficiary Household Income 
Comparatively, the degrees to which beneficiary households rely on crop farming/sales seem 
to be higher than baseline value of 66%. Following crop farming, are petty trading and casual 
labor. Overall, petty trading increased from 4.0% at baseline to 5.7% at endline, while casual 
labor reduced from 9% at baseline to 2.9% at endline. Other primary sources of beneficiary 
household income include begging, mining, charcoal, etc. Findings could be attributed to the 
fact that cash received by participant beneficiaries has given households the opportunity to 
work on their own plots and produce for themselves, rather than having to go out and look for 
temporary casual labor to get by.
“In FGDs conducted with women and men in all randomly selected locations across FEEDS 
intervention chiefdoms in Kailahun district revealed crop farming (production/sales) as their 
main occupation and source of household income. Followed by crop farming were petty 
trading and casual labour, mentioned in particularly Folu, Woloma and Baoma communities 
in Jawei, Peje West, and Upper Bambara Chiefdoms.”
“In FGDs conducted with men and women in Pendembu Njaigbla and Gbaima Koboihun 
communities in the Njaluahun Chiefdom, and Manowa and Golahun communities in the Peje 
Bongray Chiefdom, participants emphasized that the main activity that they do in these 
communities to have their daily living is farming. These included; rice production, garden work, 
oil palm production, etc.” 
Secondary Income Sources: The evaluation findings revealed evidence of casual labor, 
petty trade, and livestock farming apart from crop farming/sales as secondary sources of 
household income. On average, Figure 6 reveals marked improvement in the reduction of 
casual labor, reducing from 35% at baseline to 21.3% at endline, while livestock farming 
increases from 4% at baseline to 12.2% at endline. Petty trading relatively remains the same 
as a secondary source of income among beneficiary households. Both begging and 
remittances have been reduced from 2% at baseline to 0.3% at endline. One possible 
explanation for these findings could be the fact that cash received by household beneficiaries 
has allowed households to reduce their reliance on temporary casual labor and begging and 
25
remittances, while focusing on engaging themselves in livestock and crop farming on their 
own.
Figure 6: % of Secondary Sources of Beneficiary Household Income
Although casual labor was mentioned, “When asked about secondary income sources in 
separate FGDs with FEEDS men and women beneficiaries, participants in all the discussions 
across the randomly selected FGD locations made much mentioned of petty trading and 
livestock farming among others.” 
Tertiary Income Sources: Figure 7a shows that crop farming takes the lead for tertiary 
beneficiary income sources, followed by casual labor, others, and petty trading. Casual labor
specifically grew from 22% at baseline to 28.5% at endline. While petty trading remained
relatively the same with a baseline value of 14%, begging has largely reduced from 12% at 
baseline to 0% at endline. These are complemented by reduced utilization of other diversified 
sources of income, including charcoal sales and cash for work. A possible explanation for the 
above evidence could be the fact that with FEEDS cash, households that previously were 
engaged in begging, charcoal sales, or had no income source have increased rates of
temporary casual labor by providing their services to households engaged in farming and 
trading, thereby increasing farming and petty trading activities among FEEDS-supported 
households.
26
Figure 7a: % of Tertiary Sources of Beneficiary Household Income
Overall Sources of household income: As seen in Figure 7b, the average reliance of 
households on crop farming/sales, casual labor, petty trading, and livestock farming as the 
main sources of income has increased. While petty trading and livestock farming marginally 
increased by 0.7% and 3.1% respectively, crop farming increased largely from the baseline at 
32% to an endline value of 58.9%. Concurrently, evidence revealed that casual labor markedly 
reduced from 22% at baseline to 16% at endline, while other sources of income (begging, 
firewood and charcoal sales, cash for work, or no income) largely reduced from 18.7% to 6.3% 
at endline.
27
Figure 7b: % of Overall Sources of Beneficiary Household Income
Linking the marked improvement in the degree to which households depend on each of the 
primary, secondary, and tertiary income sources presented above, there is likely possibility for 
the evidence to remain associated with contributions from the SC FEEDS cash in the 
intervention district, causing positive changes in sources of income. As they can be seen to 
spend a little more on food and agriculture (see Figure 8 below), FEEDS-supported 
households seem to have eventually been able to have access to better own production and 
purchase of food from the market in other to meet their food security and nutritional needs.
3.2.2 Normal Household Expenditure Pattern
Figure 8 presents the orderly magnitude of average normal household expenditure patterns 
on food and non-food items. Overall, average expenditure on food is seen to have increased 
from 44% at baseline to 45.0% at endline. Figure 8 also reveals that average expenditures on 
non-food items have increased since baseline. In particular, expenditures on education moved
from 10% at baseline to 18.3% at endline, expenditure on health care increased from 8% at 
baseline to 14.7% at endline, expenditures on agriculture went from 7% at baseline to 12% at 
endline, and expenditure on savings at endline was 4.7% compared to 0% at baseline. Debt 
repayment and transportation reduced from baseline values of 11% and 5% down to 1.7% 
and 2.7% respectively. Expenses on other items (including household items, rent/shelter, gifts, 
etc) have also reduced from 15% at baseline to 1% at endline. The reduction in expenditure 
on debt repayments and other items could be possibly explained due to households spending
funds on agriculture, food, or savings to improve on their diet and food intake. 
28
The increases in expenditure on the above non-food items could be attributed to the fact that 
at baseline, while movement was restricted during Ebola, households were recovering from 
reduced access to their farmlands. Once the restriction was over, participants’ household 
engagement in crop farming for their own production of food has once again increased. While 
diverting excess funds available to expenditure on non-food items, beneficiary households 
continued to have a 1% increase in food expenses, which seemed to be used as part of inputs 
for a more diversified food production. 
Figure 8: % of Average Normal Household Expenditure Pattern
3.3 FEEDS Intervention Outcomes
This section presents evaluation findings on key outcomes indicators of the FEEDS 
intervention. These are given in averages, and include the Household Dietary Diversity Score 
(HDDS), Household Hunger Scale (HHS), Food Consumption Score (FCS), reduced Coping 
Strategies Index (rCSI), and the Meal Frequency.
3.3.1 Household Dietary Diversity Score (HDDS)
Household dietary diversity is defined as the number of food groups consumed by members 
of a given household within a 24 hour period. This gives an indication of the household’s 
access to food. Respondents were asked to recall all the different foods that household 
29
members had consumed within the last 24 hours prior to the survey. These different foods 
were grouped into the standard 12 food groups used in the calculation of the HDDS. The 
maximum possible score is 12, which would indicate that a household ate all the 12 different 
food groups within the recall period. 
Figure 9 presents evaluation findings relating to the average household dietary diversity score 
of the target beneficiaries. It can be seen that at baseline the average score was 4.5, which 
the FEEDS intervention aimed at increasing to 5. At the endline, the evaluation found that the 
average HDDS for all households was 6.3. 
Figure 9: Average Household Dietary Diversity Score (HDDS)
Results from an Independent Sample T-test of statistical significance conducted on average 
household dietary diversity score revealed a statistically significant (p<0.001) difference 
between baseline and endline values with respect to average HDDS. This is because the 1-
tailed significance value (p<0.001) is far less than the significance level 0.05 measured at 95% 
confidence level. See appendix A.
These results show a good achievement relative to the intended target. From a review of PDM 
reports, program field staff intermittently conducted nutrition sensitive sensitizations and 
awareness-raising campaigns across all target communities. These activities may have also 
contributed to the good achievements noted on the HDDS. 
The different kinds of food groups that each beneficiary household consumed in the last 24
hours prior to including them into the survey were analyzed. Figure 10 reveals the proportions 
at which beneficiaries consumed the following food groups at endline: cereals at 97.1%, fish 
at 97.2%, and roots and tubers at 90.8%. These are followed by more than 77% of fruits and 
vegetables each, and a minimum of 41% for both condiments and pulses. Over one-third of 
participants reported to have consumed fats and oils as well as sugar or honey. 
Other food groups that had limited consumption among all households included meat, poultry 
and other organ meats, dairy products, and eggs. From KIIs with CTC members and SC 
program staff, it was revealed that SC field staff intermittently conducted nutrition-sensitive 
messaging in the target communities. This finding was confirmed by FGDs with participants, 
who reported that these sensitizations played a role in helping them diversify their foods, as 
seen in the values reported above.
30
Figure 10: % of Respondents that consumed each Household Dietary Diversity Score food 
group (Multiple Responses)
In a similar manner, Chi-Square tests conducted on all individual food groups revealed that 
the difference between baseline and endline values with respect to all other individual food 
groups are statistically significant (p≤0.002), except for condiments and fat/oil, for which 
results showed existence of statistically insignificant differences between the baseline and 
endline for the two groups. See appendix A on results from statistical tests of significance. The
test results, apart from the exceptions of condiments and fat/oil, are considered statistically 
significant because the individual p-values are each far less than the significance level (0.05).
3.3.2 Household Hunger Scale (HHS).
The Household Hunger Scale is a particular household food insecurity indicator used to 
measure household hunger in food insecure areas. This indicator was determined using 
information provided by beneficiary respondents in response to questions on the extent to 
which, in the past one month or four weeks:
 Household members had no food to eat at all in the house
 Household members go to sleep at night hungry
 Household members go a whole day and night without eating anything at all because 
there was not enough food in the house. 
31
Analysis of results from household responses to these three parameter questions is presented 
in Figure 11 below. These have been categorized into three hunger levels: ‘Severe hunger’, 
‘Moderate hunger’, and ‘Little to no hunger’ among beneficiary households.
Figure 11: % of Average Household Hunger Scale
Severe hunger among all households reduced from 1.0% at baseline to 0.5% at endline (this 
is supported by the statistical test result). A second data collection exercise would have been 
necessary to investigate why 0.5% of households remain severely hungry, but time and 
resources could not permit the evaluation team to revisit these households to further 
investigate these results. 
In addition, the average moderate hunger among beneficiary households significantly reduced
from 68% at baseline to 34.3% at endline, and on average, little to no hunger showed marked 
increase from 32% at baseline to 65.2% at endline. The consistent reductions in both 
moderate and severe hunger resulted in the marked increase in percentages of households 
with little to no hunger, from 32.0% at baseline to endline value of 65.2%. A very limited 
proportion of beneficiary households continued to face severe hunger situations in FEEDS 
intervention target areas. The marked reduction of the proportion of beneficiary participants 
facing severe and moderate hunger (which was the basis of their selection at baseline), 
accompanied by a noticeable increase in the proportion of households with little to no hunger,
could be attributed to contributions by the FEEDS intervention, which in effect demonstrates 
value for money with regards to intervention.
These findings are supported by Chi-Square tests (presented in appendix A) conducted on 
three HHS parameters. It was revealed that the difference between baseline and endline 
values with respect to the three parameters (given below) were statistically significant because 
the significance values (p<0.001) for the parameters are each less than the significance level 
(0.05) measured at 95% confidence level:
● The extent to which household members in the past four weeks go a whole day and 
night without eating anything at all because there was not enough food in the house
● The extent to which household members in the past four weeks have no food of any 
kind to eat at all because of lack of food in the house
● The extent to which household members in the past four weeks go to sleep at night 
hungry because there was not enough food in the house. 
32
3.3.3 Household Food Consumption Score (FCS)
The Food Consumption Score (FCS) is mostly used as a proxy indicator for access to food. It 
is a weighted score based on dietary diversity, food frequency, and the nutritional importance 
of food groups consumed by households. The data for this score was collected by asking 
respondents to recall the number of days in the last 7 days preceding the survey that the 
household ate different food types. These food types were then grouped into the standard 12 
food groups for the calculation of FCS. The higher the score, the better the food security 
situation of the household. 
In the FEEDS intervention, this indicator was not monitored and so there was no baseline 
value to measure it against. Results from the endline are presented here without comparison 
to the baseline or midline surveys. As shown in Table 2, the Comprehensive Food Security 
and Vulnerability Analysis (CFSVA) report 2015, conducted by WFP, gives international 
benchmarks for poor, borderline, and acceptable food consumption scores.
Table 2: Food Consumption Score Benchmarks
Food Consumption Profile Food Consumption Score
Poor 1 – 21
Borderline › 21.5 – 35
Acceptable › 35
Source: WFP’s CFSVA Report 2015
When the evaluation consumption scores were compared with these international 
benchmarks, it was found that the overall average household score of 52.7 was far above the 
minimum acceptable consumption level of 35+ (see Figure 12). 
Figure 12: Average Household Food Consumption Score
The households that participated in the FEEDS project were selected after a rigorous poverty 
and vulnerability screening exercise for phases I and II. Save the Children used a participatory 
targeting approach to identify these beneficiaries, and community members were directly 
responsible for selecting eligible HHs as per the targeting criteria through a public and 
transparent process. In phase III, the FEEDS program used the HEA baseline and outcome 
analysis data extensively to re-structure the targeting, community selection, and beneficiary 
registration procedures. The steps in the process used for the re-targeting were geographical 
targeting to identify which communities were to participate, community ranking (identification 
by community members of HH matching the selection criteria), scoring - which implies that all 
identified HH were pre-registered and surveyed (scoring tool integrated to pre-registration 
questionnaire), selection, which is ranking and selection of beneficiary households based on 
scoring tool (higher scores), and verification of the HH information for those with high scores. 
33
One would assume that household food consumption scores were poor, since only 3.3% of all 
households in Kailahun were food secure and another 26.1% were marginally food secure 
(CFSVA, 2015). Therefore, the evaluation concluded that the FEEDS intervention has 
contributed immensely to improving the food consumption score of target households.
3.3.4 Household Reduced Coping Strategies Index (rCSI)
Another key indicator the evaluation assessed was the reduced Coping Strategies Index 
(rCSI), which measures the frequency of different strategies adopted by households to cope 
with shortages of food, and calculates a score weighted by the severity of each strategy.
In Figure 13 (below), it can be seen that at baseline, this index was 25 out of a possible index 
score of 56; for the rCSI, the lower the score the better. The FEEDS intervention aimed at 
reducing this score from 25 to 23. The endline evaluation found an average household coping 
strategies index to be 17.5. This is far below the FEEDS intervention target household coping 
strategies index of 23.0, which shows a marked achievement in reducing negative coping 
strategies for households in FEEDS intervention coverage areas (this claim is supported by 
the statistical test result given in appendix A). From KII with program staff, it was found that 
Save the Children, during their nutrition sensitizations to beneficiaries, always stressed the 
fact that households should not employ the coping strategy of starving adults so children could 
eat. Save the Children’s point here was that if adults are starving continuously to let children 
eat, there will come a time when they will not be healthy enough to provide for the children 
and the whole household will be in food security crisis. The target households may have 
adopted this message in their coping strategies, so that both adults and children always have 
something to eat. 
Figure 13: Average Household reduced Coping Strategies Index
The findings above are supported by results from a one-tailed Independent Sample T-test of 
statistical significance conducted at 5% degree of accuracy and 95% confidence level on 
average of the five categories/parameter measurement of reduced coping strategies index 
(rCSI) outlined below, revealed that there is a statistically significant (p<0.001) difference 
between baseline and endline values in respect to average rCSI. This is because the 1-tailed 
significance value (p<0.001) is be far less than the significance level (0.05). See appendix A.
● How many days did your household have to rely on less preferred foods or less 
expensive food?
● How many days did your household have to borrow food or rely on help from friends 
and relatives?
● How many days did your household have to reduce the portion size of meals?
34
● How many days did your household have to reduce the quantities of food consumed 
by adults so that young children could eat?
● How many days did your household have to reduce the number of meals eaten per 
day?
3.3.5 Average Household Meal Frequency 
Table 3 presents findings relating to the number of meals eaten by target households in the 
last 24 hours before they were included in the survey. This is another key measure of food 
access at the household level. The data was disaggregated by age to present the complete 
picture of the household meal situation in the last 24 hours before the survey. It was found 
that, on average, every household member ate more than one meal within that last 24 hours 
prior to the survey. Meal frequency was not computed at baseline. However, noting the criteria 
used to select the target beneficiaries, and taking into consideration the results of the reduced 
coping strategies index measured at the baseline and throughout the program, it is known that 
some households reduced the frequency of meals, at least for specific household members, 
in order to cope with reduced access to food. At endline, findings suggest that every household 
had a secured one meal plus within the last 24hrs before the survey.
Table 3: Meal Frequency (for last 24hrs)
All Total
Children less than 5yrs 1.7
Children 5-17yrs 1.8
Adults 18yrs and above 1.7
Overall Total 1.7
3.3.6 Utilization of FEEDS Cash by Beneficiaries
The evaluation found that a total of 91,154 unconditional cash transfer participants were 
targeted by the FEEDS project, distributing a total cash amount of approximately $4.9 million. 
Figure 14 presents the orderly magnitude of the average proportion of FEEDS cash 
expenditure patterns on food and non-food items. Overall, 44.3% of cash received was spent 
on food, followed by education, health care, agriculture, and savings at 18%, 15%, 12% and 
5.7% respectively. This stresses the fact that in vulnerable communities like those targeted by 
the FEEDS project, households spend a quite a large portion of their income on food. 
However, the FEEDS cash may have given them some flexibility to spend on non-food items. 
Although food remains the major item for household expenditure, the above expenditure 
pattern of FEEDS beneficiaries is consistent with the theoretical fact that once income of 
households increases, members are prompted to also prioritize other expenditure lines that 
are important for their health and welfare. These in effect account for strengthening vulnerable 
households’ human and financial capital, which are necessary for future sustainability of the 
FEEDS intervention. 
35
Figure 14: % of FEEDS Cash Utilized on Food and Non-Food Items
Types of Foods Purchased with FEEDS Cash Figure 15 presents findings related to the 
different food items on which households spent the cash received from FEEDS support. On a 
whole, it can be seen that in all phases, a minimum of 90% of beneficiary households spend 
the cash on rice, fish, vegetable oils/palm oil, and condiments. Other key food items bought 
frequently were tubers/roots (cassava, potatoes, etc.), sugar/honey, and pulses (beans, pears, 
lentils, etc.). Comparatively, vegetables, meat, fruits, and eggs seem to attract less attention 
for purchase by households. This is because these food items are generally produced locally 
by the households themselves. 
36
Figure 15: % of Food types Purchased with FEEDS Cash (Multiple Responses)
3.4 Capacity Supports to FEEDS Cash Beneficiaries 
3.4.1 Capacity Supports to Traders
The evaluation conducted several Focus Group Discussions (FGDs) with traders, primarily 
women, who were supported to do petty trading in their communities. The evaluation found 
that the FEEDS project supported these women with direct cash and capacity building training 
in Business/Financial management. Findings from the FGDs are summarized below.
● Cash Support
The evaluation found that all the female traders who participated in the FGDs reported that 
they were given cash by the FEEDS project to do business. Most of them were already doing 
petty trade in their various communities (though according to them their capital had drastically 
dwindled) and some had gone out of business due to the Ebola Virus Disease (EVD) outbreak. 
All of them reported that they were given two tranches of cash to support their businesses. 
“Yes, we were given money by Save the Children to start our businesses again after we lost 
all the little monies we had during the EVD.” [FGD Participant, Baiwala Community, Mandu 
Chiefdom] 
37
This evaluation therefore concludes that the FEEDS intervention supported female traders 
with cash to continue or restart their businesses. A total of approximately $137,878 was 
disbursed to 400 traders (399 female and 1 male) in two tranches across the project target 
communities. 
● Capacity Building Trainings
The evaluation also examined whether the female traders were given any form of capacity 
building training in business/financial management. Findings suggest that most (not all) of the 
traders who participated in the FGDs had received training from Save the Children in business 
and financial management. According to most of the traders, these trainings were conducted 
before they were given the cash, although for some the trainings were conducted after the 
cash was given to them. Whichever way it was done, results found that the duration between
the receipt of cash and training was not long enough to affect the activity intent either by using 
the cash incorrectly before the training came or forgetting the training before the cash came. 
“Because of the training given to us by Save the Children, we have now got some business 
techniques like writing report on all our transactions” [FGD Participant, Kambama Community, 
Jawei Chiefdom]
“The training has taught us to be able to engage in the right business good for us” [FGD 
Participant, Manowa Community, Peje Bongrey Chiefdom] 
A total of 400 traders were provided conditional cash grants and trainings were held across 
the target communities. 
● Impact of Cash and Capacity Building Training on Trade
The evaluation assessed the opinion of traders about how the cash and capacity building 
training affected their trade volumes and subsequently their profit margins. It found that all 
(100%) of traders reported that the cash and training support boosted their trade volumes and, 
therefore, profit margins. Traders reported that once they started receiving the cash from the 
FEEDS project, they were able to increase their stocking capacity of basic food stuff. Some 
even reported having used part of the cash to improve their storage facilities, in order to 
accommodate the increments in their stock. During implementation, the program measured 
the value of incremental sales recorded by traders and found that, within 6 to 9 months, sales 
had increased by $248,839, implying that for every $1 received $2 was generated.
However, the evaluation was not able to quantitatively confirm these claims because there 
was no monitoring data that had been tracking changes in the traders’ stock level over time 
after this initial survey. However, findings from FGDs with participants who were not traders, 
revealed that most of their basic foodstuff needs were met by their community markets, which 
was not the case before the FEEDS intervention according to the baseline findings. 
Most traders also reported that the FEEDS intervention (unconditional cash transfer to 
households) increased demand for basic food stuff in their communities. Since traders were 
also simultaneously capacitated, most reported that they now not only buy commodities in 
their towns (when vendors bring them), but also go out to periodic markets outside their 
communities to buy goods and restock. According to most, this helped to increase their profit 
margins further and made most required basic food stuffs in their local markets available. 
“The cash support has also helped us go out and buy goods to sell in our community here, 
which people now buy.” [FGD Participant, Manowa Community, Peje Bongrey Chiefdom]
“It has also help us to buy in large quantity now. Before this time, we were unable to buy goods 
in large quantity because of limited money. But with the intervention of Save the Children, we 
38
now buy goods in large quantity to sell in this community. We say thanks to Save the Children.” 
[FGD Participant, Kambama Community, Jawei Chiefdom]
“The demand for basic food stuff has increased because all what people were going out to 
buy are now with us here in the community.” [FGD Participant, Gbama Koboihun Community, 
Njaluahun Chiefdom]
“Demand has now increased significantly. They now come to buy and not to loan from us.” 
[FGD Participant, Nyanyahun Community, Upper Bambara Chiefdom] 
The evaluation concluded that the FEEDS support to female traders helped restore their 
businesses (and for some, even expanded it). This had a positive ripple effect on the 
communities in terms of availability of basic food commodities in most local markets.
3.4.2 Capacity Supports to Women in Agriculture Groups (WAGs)
Like the support to female traders, the evaluation conducted several Focus Group Discussions 
(FGDs) with members of the WAGs to assess the support given to them and to see how it 
helped them in their agricultural activities. The evaluation found that the FEEDS project 
supported these women with direct cash and capacity building training in organizational
management, improved agricultural techniques, and basic business skills. Findings from the 
FGDs are summarized below. 
● Capacity Building Trainings
The evaluation found that all the WAG members who participated in the FGDs were jointly 
trained by Save the Children and MAFFS Kailahun District. According to the participants, some 
of the topics covered included:
● Farming as a business
● Development of business plans
● Conflict management
● Leadership principles
● Record keeping
● Group formation and the different responsibilities of leaders
The above topics were consistent with the planned training topics except for one: the 
component that was supposed to deal with “improved agricultural techniques”, meaning the 
focus of the capacity building was on the business aspect of agriculture (farming as a 
business). 
● Cash Support
The evaluation found that the FEEDS intervention gave out cash to the members of the 
eighteen WAGs across the target chiefdoms in Kailahun district. According to participants, 
they used the cash to buy agricultural inputs, especially seed and payment for labor. Few of 
them reported having used the money on some non-agricultural items like school fees and 
medication. 
“Yes, we received cash from Save the Children to do our farm work. We used the money to 
buy seeds, fertilizer and food for the workers.” [FGD Participant, Baoma Community, Luawa 
Chiefdom]
39
“Save the Children gave us a specific amount and told us to use it for our farming activities to 
expand the size of land we cultivate. We thank them very much.” [FGD Participant, Kambama 
Community, Jawei Chiefdom]
● Impact of Cash and Capacity Building Training on Cultivation
All the participants said the cash and training support helped them to do better farming than 
they were doing before. On further examination, they some said this was the first time their 
households had been able to cultivate all the land that belonged to them.
“Before now, we were just cultivating part of our land, because we don’t have money to pay 
labour and buy enough seeds. But after we received the cash from Save the Children, we 
cultivated all the land.” [FGD Participant, Golahun Community, Peje Bongrey Chiefdom]
Others said the cash support also helped them to get access to quality rice and vegetable 
seeds, which they planted. For some other participants, they said the knowledge they gained 
has helped their group become more united than before.
“With the training, we all now know our roles and responsibilities in the group. So, there is 
peace now and we have focused on our objectives.” [FGD Participant, Pendembu Njaigbla 
Community, Njaluahun Chiefdom]
As part of the impact of the FEEDS support to the WAGs, some participants reported that they 
have now become role models in their communities that other women are copying from.
“We have become role models to other women who are now trying to form their own women’s 
groups.” [FGD Participant, Baiwala Community, Mandu Chiefdom]
The evaluation therefore concludes that the cash and capacity building training support given 
to the WAGs had a positive impact on the operations of their groups, as well as individually. 
Additionally, there has been some spillover effect on the community, which has been 
motivated to start more women groups.
3.5 Effectiveness of Implementation - Cash Transfer Processes
Cash interventions are becoming increasingly popular. As such, it is necessary that an 
assessment be made at the end of an intervention of not only the food security indicators that 
the program sought to address, but also the process that led to the changes observed in the 
indicators. The evaluation therefore looked at four key aspects of the cash transfer process,
and the findings are presented below. 
● Satisfaction with Distance Travelled to Pay Points
The evaluation attempted to understand how the beneficiaries felt about the distances they 
travelled to access the various pay points. Table 4 shows that 88.8% of the beneficiaries 
reported they were satisfied with the distances they covered to access their pay points. The 
table also shows that 54.7% of the beneficiaries walked less than 1 mile to access their pay 
points, while 27.4% walked between 1 and 2 miles. Only a small proportion (3.8%) of 
beneficiaries reported that they walked above 3 miles to access their pay points. This was a 
reasonable distance to cover to a pay point, taking cognizance of the target terrains, which 
means the FEEDS project positioned enough pay points to avoid beneficiaries being forced to 
walk long distances. 
“The distance between the paying points to the communities is around 2 kilometres on 
average, which is less than 40 minutes walking distance.” [KII with town chief, Baiwala].
40
Table 4: Satisfaction with Distance Travelled to Pay Points
Total (n=653) %
How far do you travel to access the FEEDS pay point?
Less than 1miles 357 54.7
1miles - 2miles 179 27.4
2.1miles - 3miles 92 14.1
Above 3miles 25 3.8
Are you satisfied with the distance travelled to the pay point?
No 73 11.2
Yes 580 88.8
● Adequacy of Cash Transferred
The evaluation assessed the perception of beneficiaries regarding the adequacy of the cash 
amount the FEEDS project was giving them to cover their food need for three months. Table 
5 shows that 62.3% of all beneficiaries held the perception that the cash given to them was 
not adequate to cover all their food needs for a three month period,2
leaving only 37.7% who 
reported that the cash they received was adequate for their food needs. 
The reasons given by most households for holding the view that the cash amount was not 
adequate to cover their three month food needs were large family size and the rapid increase 
in food stuff prices during the cash transfer period. However, the evaluation found that almost 
all households were also spending the cash received on several other things including health, 
education, transportation, and agriculture, meaning the amount left to be spent on food would 
be lower than expected. Additionally, it was noted there was some misunderstanding on the 
part of beneficiaries, because the FEEDS intention was that the cash was meant to cover only 
40% of the household food needs, not all their food needs for the three months period. 
Table 5: Adequacy of Beneficiary Cash Amount
Do you consider the amount of cash you receive adequate for your household food 
needs for three months?
Frequency %
Yes 246 37.7
No 407 62.3
Total 653 100.0
● Satisfaction with Waiting Time to Receive Cash
The evaluation also examined the opinion of beneficiaries on the time taken at the pay point 
to receive their cash. Table 6 shows that almost all beneficiaries (95.7%) reported that they 
were satisfied with the length of time taken at the pay points to receive cash. It was found that 
42.3% of all beneficiaries were spending less than 30 minutes at pay points to receive cash,
while 27.7% spent between 30 to 59 minutes and 25% spent between 1 to 2 hours. Only a 
small fraction of beneficiaries (5.1%) reported that they were waiting more than for 2 hours or 
 
2 Outside of social safety net transfers, transfers were intended to cover 40% of a households 
monthly food needs, distributed in tranches of two or three months of assistance at a time. Therefore 
it is reasonable that the cash did not meet all household food needs, but further information would be 
required to understand this finding. 
41
more at pay points before they received their cash. This indicates that the activities that lead 
to the actual transfer of cash from SPLASH to the beneficiaries was fast enough to avoid 
significant delays at the pay point.
Table 6: Satisfaction with Waiting Time to Receive Cash
Total (n=653) %
How long do you wait to receive the cash pay point?
Less than 30 min 276 42.3
30 min - 59 min 181 27.7
1hr - 2hrs 163 25.0
2 hrs. and above 33 5.1
Are you satisfied with the waiting time for you to receive the cash at the 
pay point?
No 28 4.3
Yes 625 95.7
● Satisfaction with Safety at Pay Points
Cash is attractive to all, including criminals. The evaluation therefore investigated the views of 
beneficiaries regarding their safety when they were at the pay points to collect cash, as well 
as their safety after leaving. The evaluation also revealed beneficiary’s opinions regarding any 
pressure from other sources, including local authorities, politicians, or CTC. Table 14 shows 
that 97.4% of all beneficiaries reported that they felt safe when they came to collect their cash 
at the pay points. After leaving the pay point with their cash, the table shows that almost all of 
them (95.1%) still felt safe as they returned to their respective communities. Regarding 
pressure or harassments from local leaders or CTC members to give them part of their cash, 
results indicate that this almost never happened (99.4%) as reported by beneficiaries. These 
were good indications of safety. 
“We feel safe. The payment is normally done at the town barray which is located at the center
of town. This makes it safe for all of us.” [FGD with beneficiaries, Pendembu Njeigbla].
Table 7: Satisfaction with Safety at Pay Points
Total (n=653) %
When you are at the pay point, do you feel safe?
No 17 2.6
Yes 636 97.4
After leaving the pay point with cash, do you still feel safe?
No 32 4.9
Yes 621 95.1
Have you ever been pressured by local authorities (Chiefs, Councilors); CICs, 
extended family or community members to give money you did not want to give?
No 649 99.4
Yes 4 0.6
3.6 Appropriateness of Project Design
42
The FEEDS project was designed to primarily be a direct unconditional cash support project 
to vulnerable food insecure households affected by EVD in the Kailahun district. The overall 
objective was to increase the speed at which these households would recover from the shock 
of EVD. The strategy was to ensure that all targeted households had continued access to 
nutritious food, especially throughout the lean season, so that they could reduce recourse to 
negative coping strategies. The evaluation noted that these design strategies were all 
appropriate in enhancing the achievement of project results. 
Cash transfer projects are cost effective to implement since they do not require too many 
program activities, and the transport and logistics costs associated with other alternatives are 
eliminated. Additionally, cash transfers (especially when made unrestricted) allow 
beneficiaries a great deal of latitude to choose what to buy. This brings about some sense of 
dignity and control over household expenditures, which is a basic duty of a household head. 
Thirdly, the design also allows for greater nutrition-sensitive messaging and messaging on 
how to utilize the cash received to improve their food security situation and to cope with any 
future shock. To this end, the staff were based directly in the chiefdoms to implement the
activities associated with this design aspect. 
Though the project was largely based on unconditional cash transfers, there were two 
conditional complementary activities implemented during the program: conditional cash grants 
to traders (primarily female) and conditional transfers to Women in Agriculture Groups 
(WAGs). To receive cash assistance, these beneficiaries were required to attend trainings and 
demonstrate application of the skills and methods. This design aspect was very apt because 
it emphasizes the important aspect that food security can only be sustainably realized through 
enhancing a households or a nations own food production. 
3.7 Intervention Relevance and Strategic Fit
The Ebola Virus Disease (EVD) entered Sierra Leone through Kailahun district. From all 
indications, it took the government unawares, with no response plan in place, which is why 
the virus killed many people in Kailahun district and stayed there for a longer period than most 
other parts of the country that it affected. It was here that government and its emergency 
response partners first imposed quarantines. Whole communities and sometime whole 
chiefdoms were put under quarantine, disrupting farming activities, markets, schools, and all 
other social activities. This had a prolonged negative impact on the livelihoods of people in 
Kailahun district, and created a shock that was reported in the baseline report as follows:
“The Prevalence of households with moderate or severe hunger in Kailahun district is 69%”. 
The Reduced Coping Strategies Index score in Kailahun district is 25.” 
These two indicators showed that there was a high level of food insecurity in the district at the 
time the FEEDS baseline was conducted. Therefore, an intervention aimed at improving 
continuous access to nutritious foods by target households was relevant to Kailahun district. 
The beneficiaries themselves echoed the relevance of FEEDS project; almost all of the 
beneficiaries said if the FEEDS project had not come to their aid after EVD, they would have 
coped with life in an extremely difficult manner, which highlights the relevance of the FEEDS 
project to the target beneficiaries. 
The current strategic plan of Save the Children in Sierra Leone sets out its objectives in five 
thematic areas: Health and Nutrition, Education, Child Poverty, Child Protection, and Child 
Rights Governance. The evaluation found that the FEEDS project strategically fit into the first 
thematic area, Health and Nutrition. The Strategic Goal 1 and Strategic Objective 1 of USAID 
states, “Strengthen America’s Economic Reach and Positive Economic Impact” and “Promote 
Inclusive Economic Growth, Reduce Extreme Poverty and Improve Food Security.” The 
43
evaluation found that the FEEDS project fit into these two strategic guiding statements of the 
donor, USAID. 
Following the end of EVD, the government of Sierra Leone drew up its National Ebola 
Recovery Plan (2015 – 2017) to guide and coordinate all recovery efforts. One of the key 
objectives of the plan was, “Managing and Mitigating Immediate Ebola Impact in the Social 
Sector, Gender, Children, and Social Protection”. In order to achieve the objectives of the 
recovery plan, the government formulated certain strategies, including “Provide livelihood 
support for EVD affected children, women, orphans, widows, widowers, and the elderly and 
disabled” and “Provide cash transfers to poor households and vulnerable groups, therefore 
benefitting local economies.”
The FEEDS project undertook cash transfers to vulnerable EVD affected households and 
supported women to do business and agriculture (which are livelihood activities). The 
evaluation therefore concluded that the FEEDS project also directly fit into the government of 
Sierra Leone’s strategic plan for recovering from the shock of EVD, thus making it relevant at 
national level. 
3.8 Resource Mobilization and Efficiency of FEEDS Implementation
The evaluation found that all three categories of resources (human, material, and financial)
were mobilized within the required time frame and were adequate. The FEEDS project 
recruited staff for all required positions; field staff were based in all targeted chiefdoms and 
support staff were based in the Kailahun field office. The field staff were in direct contact with 
beneficiaries on a day-to-day basis to implement and monitor project activities, such as market 
prices, cash usage, and supporting MAFFS in conducting improved agriculture trainings. The 
field office in Kailahun provided backstopping support to the field staff in all their activities,
including liaising with MAFFS and the country office in Freetown. Because there was adequate 
staff for the required project activities, the evaluation did not find instances of staff being over￾stretched to cover large areas of implementation, or moved from one chiefdom to the other to 
support because of work overload. The evaluation believes that the human resource of the 
FEEDS project was wisely utilized.
The evaluation did not look at budgets and expenditure documents because of limited time. 
However, a review of the activity implementation against planned timelines revealed that 
almost all activities were implemented within the planned timeline, and there was no instance 
of activity backlogs due to lack of funds. All staff were paid their remunerations on time. All 
planned cash transfers (across all three phases) were completed and the required 
adjustments due to price changes were all fully accommodated without a break. The 
evaluation noted that there were no activities implemented outside the planned budgets,
indicating that FEEDS project finances were wisely utilized to achieve the desired results. 
Materials resources mobilization and utilization were also briefly assessed. The evaluation did 
not look at the project assets register, but found that adequate materials were mobilized to 
facilitate implementation: all staff required to use computers had them, motor bikes were 
pooled for field trips, and the Kailahun field office had a photocopier and printers assigned to 
the FEEDS project. At the time of evaluation, all of this equipment were in good working order. 
Regarding alternative approaches to get the same results, it was found that an online money 
transfer would have been better and faster. However, the availability of reliable mobile 
networks to support mobile transfers was a problem. Additionally, the number of target 
households with mobile phones was very low (6%) at baseline, meaning overall technology 
literacy was low. Therefore, the approach used was the most feasible under the 
circumstances. It is the conviction of the evaluation team that the FEEDS project materials 
were adequate and were utilized efficiently in the implementation of the project.
44
3.9 Partnerships, Collaborations and Coordination
The evaluation assessed how Save the Children related with other state and non-state actors 
in the implementation of the FEEDS project. This section presents the findings.
The evaluation noted that Save the Children was the sole implementing partner of the FEEDS 
project, but they collaborated with several agencies to increase efficiency and effectiveness.
The National Commission for Social Action (NaCSA) was brought on board to support 
beneficiary targeting at the community level. This was because NaCSA has a long-standing 
reputation for supporting vulnerable households through cash transfers and other 
interventions. In phases I and II, SC conducted registration using vulnerability criteria 
determined with communities, and in consultation with NaCSA, in order to select the most 
vulnerable Ebola-affected households. However, in phase III, Save the Children used the HEA 
to select target beneficiaries. NaCSA also trained Save the Children staff in the appropriate 
use of this tool. This was a capacity building activity conducted by NaCSA. In addition, the 
evaluation noted NaCSA also intermittently participated in joint monitoring activities of the 
project. 
Cash handling and transfer was outsourced to a leading mobile money transfer agency, 
SPLASH Mobile Money Limited. SPLASH specializes in both electronic and manual money 
transfer services and has operated in Sierra Leone for quite some time. Their experience with 
the local terrain was heavily leveraged upon and enabled the necessary speed and accuracy
to the cash transfer process. Their effectiveness was confirmed by beneficiaries in focus group 
discussions, most of whom said they waited for less than 30 minutes at pay points before 
receiving cash. The evaluation also confirmed from beneficiaries that they had not been 
experiencing shortages of cash transferred to them, because “the SPLASH and Save the 
Children staff always tell us to count our cash before leaving the pay point.”
The Anti-Corruption Commission (ACC), the government agency responsible for fighting 
corruption in Sierra Leone, collaborated with Save the Children to address issues of 
complaints (from both beneficiaries and non-beneficiaries) and to monitor the cash transfer 
process across the board. The evaluation found that the ACC carried out a wide range of 
sensitizations on corruption issues, its dangers, and how/why to prevent it. The evaluation 
noted they were sometimes present at cash distribution points to get first-hand information 
from beneficiaries. This was a very good strategy to add integrity to the process, and ACC has 
built a strong reputation for fighting corruption over the years despite their own challenges. 
The Ministry of Agriculture Forestry and Food Security (MAFFS), Kailahun district was brought 
on board to support the Women in Agricultural Groups (WAGs) with a series of trainings in 
improved agricultural techniques. The evaluation noted that these trainings, conducted jointly 
by MAFFS and Save the Children, enhanced the self-food production capacities of the WAGs 
members and were noted to be one of the two key sustainability activities incorporated into 
the project design.
Other key actors that Save the Children collaborated with included the Kailahun District Local 
Council, the Ministry of Social Welfare, Gender and Children’s Affairs, and the Presidential 
Recovery Team Kailahun district. The evaluation found that these other agencies were 
involved with intermittent joint monitoring on the project.
Regarding linkages of the FEEDS project with other Save the Children projects, the evaluation 
found that the nutrition-sensitive awareness raising messages that were sent out integrated 
with health, child safeguarding, education, and child rights governance messages, thus 
making the FEEDS project interface with all the other Save the Children thematic areas. 
However, there was limited involvement of the private sector, as only SPLASH was involved.
45
3.10 Mainstreaming Gender and Sustainability in FEEDS Project Activities
● Gender Mainstreaming
The evaluation assessed how gender was integrated into the project design and how that 
played out on the ground during implementation. 
The evaluation found that, at design stage, key gender issues were incorporated into the 
project. The design made way for a minimum of 50% of all beneficiaries to be women. In phase 
I, the design made way for targeting traders (mainly females) with cash support and capacity￾building trainings to continue and even expand their businesses. In phase III, the design 
targeted Women in Agriculture groups with cash and capacity building trainings to improve 
their agricultural activities. In these ways, the design allowed most of the FEEDS support, 
especially the cash for the household, to be put into the hands of more women than men. 
The project did not have a dedicated gender focal person at the field level in Kailahun. 
However, the gender coordinator based in the country office in Freetown provided relevant 
trainings and intermittent field visits during the course of implementation. Additionally, the 
program staff who were assigned to different chiefdoms also sensitized the target households 
about the benefits of joint decision-making regarding the FEEDS cash utilization, an aspect of 
gender equality that probably helped in mitigating intra-household conflicts. 
Figure 16, below, presents findings relating to how decisions were made about how to spend 
the FEEDS cash. Most decisions about how to spend the cash were made solely by those 
who received the cash themselves (67.2%). Even though this was against the notion of joint 
decision making by key household members (preferably the husband and wife), it was good 
for female beneficiaries as it means most of them had the authority to decide on how to spend 
the money, which was a form of women empowerment. Nevertheless, joint decision making 
about how to spend the cash was observed in an appreciable number of households (19.3%),
compared to decisions made by head of households alone. 
Figure 16: % of Household Member Participation in Decision Making
● Sustainability of Project Activities/Achievement
The FEEDS project main activity was direct cash transfers to vulnerable, food insecure 
households. The evaluation did not find any current measures taken by any organization or 
agency to sustain this, even though there are plans by Save the Children to handover the 
FEEDS beneficiaries to NaCSA. However, NaCSA currently operates in only 3 of the 14 
chiefdoms in Kailahun district, and an immediate scaling up by NaCSA is not in sight. There 
are also plans to hand over the WAGs to MAFFS for continuation of project activities. Again, 
46
MAFFS capability to sustain these activities (conducting trainings, providing regular extension 
services at their own expense) is not currently available. 
However, the complimentary activities provided a good sustainability framework. The 
evaluation found that, through the support of Save the Children and the other stakeholders 
(USAID, MAFFS etc.), beneficiaries have made investments in petty trading, VSL, and 
increments in land areas currently cultivated. With their enhanced knowledge of organization 
management including conflict management, roles and responsibilities, record keeping, 
business planning, and good agronomic practices skills such as nursery management, 
transplanting, spacing, weed management, integrated pest management, soil and water 
management, harvesting and post-harvest management, beneficiaries self-food production 
capabilities will be enhanced long after the end of the FEEDS project, which will hopefully
prepare them for any other shocks in the short or long term.
3.11 Monitoring, Evaluation, Accountability and Learning
The evaluation found that the FEEDS project was monitored both internally and externally. 
Internally, the field staff assigned to the chiefdoms were responsible for first line monitoring of 
project activities. They conducted post distribution monitoring (PDMs) exercises after each 
phase of cash distribution. These were meant to get feedback from beneficiaries about several 
issues, including their satisfaction with the adequacy of cash amount given to them to cover 
their food needs for three months, the waiting time at pay points before receiving cash, 
distances covered to access the pay point, cash utilization to ensure they were spending more 
on food, etc. The evaluation found that these monitoring exercises were conducted regularly 
and feedback was shared with the Kailahun field office, which in turn shared it with the MEAL 
team in the Freetown country office. Staff also conducted regular monthly market price 
monitoring of basic foodstuffs to ensure the purchasing power of beneficiaries was not 
seriously affected by price increases. This data was also shared with the respective offices for 
detail analysis. The program staff based in the field office in Kailahun, and the MEAL team in 
Freetown, provided direct monitoring support to the field staff. The Save the Children MEAL 
team based in the USA also provided remote monitoring support from time to time. The 
evaluation noted that this approach made internal monitoring more robust. 
At the external level, the evaluation found that the FEEDS project had several monitoring 
mechanisms. The first mechanism was the Community Feedback and Response Mechanism 
(CFRM). This was set to provide community members (beneficiaries and non-beneficiaries),
including children, with access to a safe and confidential means of passing feedback and 
complaints on issues relating to the FEEDS implementation. Another monitoring mechanism 
employed for external monitoring of the project was the ACC Grievance Redress Mechanism 
(GRM). This mechanism allowed beneficiaries and non-beneficiaries to call directly on a toll￾free line and raise any issues of concern relating to the FEEDS project. These concerns and 
complaints were then collated and analyzed by the respective organizations and shared with 
stakeholders for decision-making. The evaluation assessed the awareness level of 
beneficiaries about these feedback mechanisms. 
Table 8 below shows that 89.1% of beneficiaries were aware of these monitoring mechanisms. 
The evaluation noted that there was an appreciable level of awareness about the feedback 
mechanisms instituted by the FEEDS team and their partners.
Regarding the channels they were mostly aware of, the evaluation results presented in Table 
8 below found telephone hotlines to be ranked the highest at 76.4%, followed by suggestion 
boxes at 42.7% and letters at 19.6%. 
Table 8: Level of Beneficiary Awareness about the Feedback Mechanisms
47
Frequency %
If you are treated badly by a cash out/field agent, are you aware of somewhere 
to complain him/her?
Yes 582 89.1
No 71 10.9
Which channel or means do you report your concerns to the above institutions or 
persons? (Multiple responses)
Suggestion Box 279 42.7
Telephone Hotlines 499 76.4
Letters 128 19.6
Others (Specify) 34 5.2
 
Other external monitoring mechanisms employed included intermittent cash transfer working 
group meetings and food security group meetings at the national level to share lessons learnt 
and best practices. These informed field-level decisions as project implementation 
progressed, and the meetings were replicated at the district level. Outcomes were shared with 
key stakeholders, including line ministries and district local council. The evaluation therefore 
concludes that the FEEDS project was adequately monitored. 
As a supplement to the routine monitoring exercises, the evaluation found that phase I 
implementation was evaluated by the FEEDS Program Team at midline. The evaluation notes 
that the midline review findings, which were not assessed in this evaluation, showed positive 
progress towards intended results and were in line with the findings of this evaluation, even 
though the midline was not an independent review. The other two phases were not evaluated 
at midline. However, the PDMs were used as proxies to keep track of progress and learn from 
what was going on in the field.
3.12 Constraining and Facilitating Factors of Project Implementation
The evaluation noted most constraining factors of the FEEDS implementation were general 
operational constraints. There weren’t specific constraints that prevented the achievement of 
results. 
● Constraining Factors
● Bad road network and poor mobile communication network slowed down the pace 
of activities, including cash transfer itself and monitoring activities.
● Some Cash Transfer Committee (CTC) members reported that in some cases they 
felt the urge to quit the project because ‘they were not getting any direct benefit 
(tokens, etc.) in return for their services.’ This was confirmed from KIIs with 
community leaders and field program staff, who reported that they sometimes had 
to spend a long time encouraging CTC members to continue working and not quit. 
According to them, committed new CTC members onboard and get them understand 
the intervention would take too much time. Because the original CTC members 
involved already understood the project concept, and partly because they already 
knew the cash beneficiaries, it was preferable to keep them on. This problem of CTC 
members quitting the project was largely mitigated by giving them FEEDS 
customized T-shirts.
● Facilitating Factors
● The FEEDS project had enough staff to execute all key functions. Assigning 
specific staff to specific chiefdoms ensured that the project team was always close 
48
to beneficiaries. This enhanced monitoring activities and flagged any issues that 
were addressed promptly.
● The evaluation noted that all the partners the FEEDS team worked with (including 
the CTC and local authorities) performed their respective duties. This enhanced 
the level of results that the evaluation found. 
Note: Though some members of the CTC sometimes complained of “no tokens”, 
their involvement played a very crucial role in achieving the overall result that the 
FEEDS project produced.
● The project was well funded. This was a huge source of timely and full 
implementation of activities and permitted the achievement of desired results.
● Effective sensitizations and house-to-house visits by field staff about the correct 
use of cash received, coupled with complementary trainings, ensured that 
beneficiaries used more of the cash to access food rather than non-food items. 
This helped in improving their food security statuses, as evidenced by changes 
in the respective indicators. The special attention paid to household decision￾making regarding the use of FEEDS cash also helped avert most potential 
conflicts. This may have been the source of the level of women empowerment 
the evaluation found. 
4.0 Lessons Learned and Recommendations
4.1 Lessons Learned
❖ When the income of households increases in cash, they are motivated to increase their 
expenditure on other important items that were not previously prioritized as expenditure 
items when there was less or no income in the household. The FEEDS cash expenditure 
patterns seem to be contrary to the expectations of the FEEDS intervention that at the 
end of the intervention, 70% of the EFSP cash received by households would be spent 
on food items. However, the observed decrease in EFSP cash expenditure on food 
(44.3%) was compensated for by expenditures in other productive categories, including 
education, health care, agriculture, and savings, which will promote long-term food 
security. 
❖ Save the Children deployed staff directly in the target chiefdoms. This provided for a 
closer link between implementing staff and the beneficiaries throughout the life of the 
project, which was very helpful for effective monitoring of market prices, cash utilization 
and nutrition-sensitive messaging. This strategy created high SC/USAID visibility, which
helped improve implementation through prompt and firsthand information gathering that 
was used for informed decision-making about the project.
❖ Establishing the right partnerships with state and non-state actors adds special value to 
quick – impact projects like cash transfer projects. The presence of NaCSA, ACC, KDC 
and CTCs added integrity and visibility to the whole process. 
❖ When the right beneficiary targeting is done, cash interventions could have a very big 
impact on the daily lives of the beneficiaries. Their food security situation will be 
significantly impacted in the shortest possible time, as measured by the statistically 
significant changes that occurred in the values of the target household’s HDDS, HHS,
and rCSI.
❖ The involvement of CTCs had a meaningful impact, especially in targeting the right 
beneficiaries and conveying information quickly to beneficiary households. However, 
when they are not allowed to benefit directly, they could have problems with total 
commitment.
❖ In the absence of effective mobile technology, a well-deployed and monitored offline cash 
transfer project can adequately replace the technology-based cash transfers. 
49
4.2 Recommendations
1. The average meal frequency of households per day was noted to be approximately 2 
at the final evaluation. This figure should reach a minimum of 3 for better nutrition
outcomes. Therefore, the Government of Sierra Leone, through the Ministry of Health 
and Sanitation and represented by the District Health Management Team (DHMT) and 
other related community structures, should continue with activity-based monitoring 
program strategies to increase or maintain the community’s access to sufficient 
nutritious foods in the right varieties.
2. The evaluation found that many households have diversified their diets as a result of 
the EFSP intervention. This has prompted most to start or increase their backyard 
gardening to keep up a steady stream of fresh vegetables, fruits, etc. On that note, the 
Government of Sierra Leone, through the Ministry of Agriculture, Forestry and Food 
Security (MAFFS), should undertake regular sensitizations on the benefits of backyard 
gardening and the importance of eating these foods instead of selling them to help 
improve household nutritional status. Additionally, MAFFS block extension officers 
should continue monitoring the participants who were trained in post-harvest 
management technology to ensure they continue to employ it in their farming activities. 
They should also provide technical assistance where necessary. 
3. The cash transfer processes and systems were largely effective, as noted by the 
satisfaction levels of beneficiaries with the distances covered to access pay points, 
waiting time to receive cash, and safety both at pay points and after leaving the pay 
points. It is recommended that the strategies employed to achieve these must be 
maintained by Save the Children in similar future interventions.
4. The direct deployment of Save the Children staff in the target chiefdom helped 
extensively in the achievement of results. This strategy should be replicated and/or 
scaled up in future similar interventions.
5. There was some misunderstanding on the part of the beneficiaries about the 
percentage of household food needs that the transferred cash was supposed to cover.
Most of them thought the cash was meant to cover all the household food needs, so
most households (62.1%) reported that the cash was not adequate for their household 
food needs. It is recommended that in the future, a very clear sensitization must 
precede the transfer of cash and this should continue throughout the life of the project.
6. The evaluation noted that there was limited involvement of the private sector in the 
overall implementation. Though there was not observed direct negative impact on the 
project, the high potential of this sector could be tapped into if more organizations from 
the sector were drawn into the project. It is recommended that in similar future projects, 
Save the Children should partner with more than one private sector organization. 
50
References 
1. Ballard, Terri; Coates, Jennifer; Swindale, Anne; and Deitchler, Megan. 2011. Household 
Hunger Scale: Indicator Definition and Measurement Guide. Washington, DC: FANTA-2 
Bridge, FHI 360
2. Bamberger, M., Rugh, J. & Mabry, L., 2012. RealWorld Evaluation: Working Under 
Budget, Time, Data, and Political Constraints. (Second edition). Thousand Oaks: Sage.
3. CARE International, 2009. Guidelines for the implementation of Baseline study for 
women’s empowerment programs funded by Norad (2009 - 2013)
4. Kennedy, Gina; Ballard, Terri; and Dop, MarieClaude. 2013. Guidelines for Measuring 
Household and Individual Dietary Diversity. EC-FAO “Linking Information and Decision 
Making to Improve Food Security” Program.
5. Krejcie and Morgan, 1970. Determining Sample Size in Research Activities. Educational 
and Psychological Measurement, No. 30, pp.607-610.
6. Magnani, R., 1999. Sampling guide. Food and Nutrition Technical Assistance Project 
(FANTA) and Academy for Educational development. The document can be downloaded 
from http://www.fantaproject.org
7. Maxwell, D., and Caldwell. R, 2008. Coping Strategies Index: Field Methods Manual. A 
Tool for Rapid Measurement of Household Food Security and the Impact of Food Aid 
Programs in Humanitarian Emergencies. Second Edition, Cooperative for Assistance 
and Relief Everywhere, Inc. (CARE).
8. Poverty Reduction Strategy Paper (PRSP) or the Agenda for Prosperity (PRSP III –
2013-2018)
9. Sierra Leone Demographic and Health Survey (2013) 
10. Sierra Leone National Ebola Recovery Plan (2015 – 2017)
11. Sierra Leone World Food Program (WFP) Comprehensive Food Security Vulnerability 
Assessment (CSFVA), 2015.
12. Stukel, Diana; and Deitchler, Megan. 2012. 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. Washington, DC: FHI 360/ FANTA.
13. Swindale, Anne, and Paula Bilinsky. 2006. Household Dietary Diversity Score (HDDS) 
for Measurement of Household Food Access: Indicator Guide (v.2). Washington, D.C.: 
FHI 360/FANTA.
14. WFP, 2008. Food Consumption Analysis: Calculation and Use of the Food Consumption 
Score in Food Security Analysis. Technical Guidance Sheet on Vulnerability Analysis 
and Mapping 
51
Appendices
Appendix A: Results from Statistical Test of Significance
Chi-Square Test Results on Household Hunger Scale (HHS)
Group * In the past one month, was there ever no food to eat of any kind in your 
house because of lack of resources to get food?
Crosstab
In the past one month, was there ever no food to 
eat of any kind in your house because of lack of 
resources to get food?
No Yes Total
Group Baseline Count 134 266 400
Expected Count 217.7 182.3 400.0
Endline Count 439 214 653
Expected Count 355.3 297.7 653.0
Total Count 573 480 1053
Expected Count 573.0 480.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 113.760a 1 .000
Continuity Correctionb
112.405 1 .000
Likelihood Ratio 115.302 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 182.34.
b. Computed only for a 2x2 table
Group * In the past one month, did you or any household member go to sleep at night 
hungry because there was not enough food?
Crosstab
In the past one month, did you or any household 
member go to sleep at night hungry because there 
was not enough food?
No Yes Total
Group Baseline Count 110 290 400
Expected Count 189.2 210.8 400.0
Endline Count 388 265 653
Expected Count 308.8 344.2 653.0
Total Count 498 555 1053
Expected Count 498.0 555.0 1053.0
52
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 101.380a 1 .000
Continuity Correctionb
100.104 1 .000
Likelihood Ratio 104.203 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 189.17.
b. Computed only for a 2x2 table
Group * In the past one month, did you or any household member go a whole day and 
night without eating anything at all because there was not enough food?
Crosstab
In the past one month, did you or any household member 
go a whole day and night without eating anything at all 
because there was not enough food?
No Yes Total
Group Baseline Count 183 217 400
Expected Count 240.8 159.2 400.0
Endline Count 451 202 653
Expected Count 393.2 259.8 653.0
Total Count 634 419 1053
Expected Count 634.0 419.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 56.286a 1 .000
Continuity Correctionb
55.317 1 .000
Likelihood Ratio 56.079 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 159.16.
b. Computed only for a 2x2 table
53
A one-tailed Independent Sample T-Test Results on Average 
Household Dietary Diversity Score (HDDS)
Group Statistics
Group N Mean Std. Deviation Std. Error Mean
Average 
HDDS
Baseline 400 4.60 2.084 .104
Endline 653 6.44 1.971 .077
Independent Samples Test
Levene's Test for 
Equality of Variances t-test for Equality of Means
F Sig. t df
Sig. (2-
tailed)
Mean 
Difference
Std. Error 
Difference
90% Confidence 
Interval of the 
Difference
Lower Upper
Average 
HDDS
Equal variances 
assumed .770 .380 -14.401 1051 .000 -1.842 .128 -2.053 -1.631
Equal variances 
not assumed -14.210 807.770 .000 -1.842 .130 -2.055 -1.629
By dividing two-tailed test significance value in half, after running the test at 90% CI will 
result into a one-tailed test significance value as presented in the body of the report. 
Chi-Square Tests Results on Individual Food Groups
Group * Corn/maize, sorghum, millet, wheat, bread, biscuits, rice, rice noodles
Crosstab
Corn/maize, sorghum, millet, wheat, 
bread, biscuits, rice, rice noodles
No Yes Total
Group Baseline Count 93 307 400
Expected Count 42.5 357.5 400.0
Endline Count 19 634 653
Expected Count 69.5 583.5 653.0
Total Count 112 941 1053
Expected Count 112.0 941.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 107.972a 1 .000
Continuity Correctionb
105.843 1 .000
Likelihood Ratio 107.926 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 42.55.
b. Computed only for a 2x2 table
54
Group * Potatoes, sweet potatoes, yams, manioc, cassava, coco yams or any other 
foods made from roots or tubers
Crosstab
Potatoes, sweet potatoes, yams, 
manioc, cassava, coco yams or any 
other foods made from roots or tubers
No Yes Total
Group Baseline Count 103 297 400
Expected Count 61.9 338.1 400.0
Endline Count 60 593 653
Expected Count 101.1 551.9 653.0
Total Count 163 890 1053
Expected Count 163.0 890.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 52.003a 1 .000
Continuity Correctionb
50.745 1 .000
Likelihood Ratio 50.438 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 61.92.
b. Computed only for a 2x2 table
Group * Vegetables like spinach, tomato, onion, okra, eggplant, pepper, or pumpkin
Crosstab
Vegetables like spinach, tomato, 
onion, okra, eggplant, pepper, or 
pumpkin
No Yes Total
Group Baseline Count 138 262 400
Expected Count 70.7 329.3 400.0
Endline Count 48 605 653
Expected Count 115.3 537.7 653.0
Total Count 186 867 1053
Expected Count 186.0 867.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 125.715a 1 .000
Continuity Correctionb
123.855 1 .000
Likelihood Ratio 123.523 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 70.66.
55
b. Computed only for a 2x2 table
Group * Fruits like mango, guava, papaya, banana, avocado, orange, lemons or 
plantains
Crosstab
Fruits like mango, guava, papaya, 
banana, avocado, orange, lemons or 
plantains
No Yes Total
Group Baseline Count 180 220 400
Expected Count 123.8 276.2 400.0
Endline Count 146 507 653
Expected Count 202.2 450.8 653.0
Total Count 326 727 1053
Expected Count 326.0 727.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 59.493a 1 .000
Continuity Correctionb
58.438 1 .000
Likelihood Ratio 58.610 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 123.84.
b. Computed only for a 2x2 table
Group * Beef, pork, lamb, goat, rabbit wild game, chicken, duck, or other birds, liver, 
kidney, heart, or other organ meats
Crosstab
Beef, pork, lamb, goat, rabbit wild 
game, chicken, duck, or other birds, 
liver, kidney, heart, or other organ 
meats
No Yes Total
Group Baseline Count 378 22 400
Expected Count 359.7 40.3 400.0
Endline Count 569 84 653
Expected Count 587.3 65.7 653.0
Total Count 947 106 1053
Expected Count 947.0 106.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 14.857a 1 .000
Continuity Correctionb
14.055 1 .000
Likelihood Ratio 16.083 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
56
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 40.27.
b. Computed only for a 2x2 table
Group * Fresh or dried fish or shellfish
Crosstab
Fresh or dried fish or shellfish
No Yes Total
Group Baseline Count 145 255 400
Expected Count 61.9 338.1 400.0
Endline Count 18 635 653
Expected Count 101.1 551.9 653.0
Total Count 163 890 1053
Expected Count 163.0 890.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 212.689a 1 .000
Continuity Correctionb
210.137 1 .000
Likelihood Ratio 218.899 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 61.92.
b. Computed only for a 2x2 table
Group * Foods made from beans, peas, lentils, or nuts
Crosstab
Foods made from beans, peas, lentils, 
or nuts
No Yes Total
Group Baseline Count 322 78 400
Expected Count 268.2 131.8 400.0
Endline Count 384 269 653
Expected Count 437.8 215.2 653.0
Total Count 706 347 1053
Expected Count 706.0 347.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 52.840a 1 .000
Continuity Correctionb
51.863 1 .000
Likelihood Ratio 55.280 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
57
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 131.81.
b. Computed only for a 2x2 table
Group * Cheese, yogurt, milk or other milk products
Crosstab
Cheese, yogurt, milk or other milk 
products
No Yes Total
Group Baseline Count 398 2 400
Expected Count 390.5 9.5 400.0
Endline Count 630 23 653
Expected Count 637.5 15.5 653.0
Total Count 1028 25 1053
Expected Count 1028.0 25.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 9.775a 1 .002
Continuity Correctionb
8.515 1 .004
Likelihood Ratio 12.145 1 .000
Fisher's Exact Test .001 .001
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 9.50.
b. Computed only for a 2x2 table
Group * Foods made with oil, fat, or butter
Crosstab
Foods made with oil, fat, or butter
No Yes Total
Group Baseline Count 260 140 400
Expected Count 249.2 150.8 400.0
Endline Count 396 257 653
Expected Count 406.8 246.2 653.0
Total Count 656 397 1053
Expected Count 656.0 397.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 2.005a 1 .157
Continuity Correctionb
1.823 1 .177
Likelihood Ratio 2.013 1 .156
58
Fisher's Exact Test .169 .088
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 150.81.
b. Computed only for a 2x2 table
Group * Eggs
Crosstab
Eggs
No Yes Total
Group Baseline Count 391 9 400
Expected Count 369.6 30.4 400.0
Endline Count 582 71 653
Expected Count 603.4 49.6 653.0
Total Count 973 80 1053
Expected Count 973.0 80.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 26.273a 1 .000
Continuity Correctionb
25.059 1 .000
Likelihood Ratio 30.982 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 30.39.
b. Computed only for a 2x2 table
Group * Sugar or honey
Crosstab
Sugar or honey
No Yes Total
Group Baseline Count 346 54 400
Expected Count 289.8 110.2 400.0
Endline Count 417 236 653
Expected Count 473.2 179.8 653.0
Total Count 763 290 1053
Expected Count 763.0 290.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square 63.719a 1 .000
59
Continuity Correctionb
62.589 1 .000
Likelihood Ratio 68.464 1 .000
Fisher's Exact Test .000 .000
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 110.16.
b. Computed only for a 2x2 table
Group * Other foods, such as condiments, coffee, tea
Crosstab
Other foods, such as condiments, 
coffee, tea
No Yes Total
Group Baseline Count 239 161 400
Expected Count 232.9 167.1 400.0
Endline Count 374 279 653
Expected Count 380.1 272.9 653.0
Total Count 613 440 1053
Expected Count 613.0 440.0 1053.0
Chi-Square Tests
Value df
Asymptotic 
Significance (2-
sided)
Exact Sig. (2-
sided)
Exact Sig. (1-
sided)
Pearson Chi-Square .625a 1 .429
Continuity Correctionb
.527 1 .468
Likelihood Ratio .626 1 .429
Fisher's Exact Test .440 .234
N of Valid Cases 1053
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 167.14.
b. Computed only for a 2x2 table
A one-tailed Independent Sample T-Test Results for Average 
Reduced Coping Strategies Index (RCSI)
Group Statistics
Group N Mean Std. Deviation Std. Error Mean
Average rCSI Baseline 400 3.206 1.3638 .0682
Endline 653 2.442 1.2778 .0500
60
Independent Samples Test
Levene's Test for 
Equality of Variances t-test for Equality of Means
F Sig. t df
Sig. (2-
tailed)
Mean 
Difference
Std. Error 
Difference
90% Confidence 
Interval of the 
Difference
Lower Upper
Average 
rCSI
Equal variances 
assumed 3.574 .059 9.181 1051 .000 .7643 .0832 .6273 .9014
Equal variances not 
assumed 9.039 801.643 .000 .7643 .0846 .6251 .9036
By dividing two-tailed test significance value in half, after running the test at 90% CI will 
result into a one-tailed test significance value as presented in the body of the report.
Appendix B: List of Evaluation Team members
No. Name Designation
1 Baimba Abdulai Koroma Team Leader (Food Security and Cash 
Transfer Evaluation Expert)
2 Momoh Thomas Bockarie Socio-economic and Livelihood Expert
3 Kenneth Lansana Mangow Field supervisor
4 Aiah John Kellie Field supervisor
5 Jesse Yamba Online Kobo Mobile-based Data Collection 
Platform System Designer
6 Joseph Kallon Enumerator
7 Mohamed Daramy Enumerator
8 Sahr Emmanuel Koademba Enumerator
9 Andrew Senesie Enumerator
10 Kenneth Maada Lansana Enumerator
11 David Foday Enumerator
12 Mohamed Gbabai Enumerator
13 Edward W. Baion Enumerator
14 Hawa Musa Enumerator
15 Haja M. Musa Enumerator
16 Mohamed Brima Enumerator
17 Manu Katta Enumerator
18 Mohamed Konneh Enumerator
19 Salieu Koroma Enumerator
20 Mamusu Gordon Enumerator