February 2017 This publication was produced at the request of the United States Agency for International Development. It was prepared independently by Melissa Chiappetta, Abigail Conrad, Claire Goydan, and Sarah Kozyn of the Rapid Feedback Monitoring, Evaluation, Research, and Learning Consortium. Keeping Families Together: A Formative Evaluation for two Social and Behavior Change Communication Campaigns for the Family Care First Cambodia Project KEEPING FAMILIES TOGETHER A FORMATIVE EVALUATION FOR TWO SOCIAL AND BEHAVIOR CHANGE COMMUNICATION CAMPAIGNS FOR THE FAMILY CARE FIRST CAMBODIA PROJECT February 2017 DISCLAIMER The author’s views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government. ACKNOWLEDGEMENTS The evaluation team would like to thank Emily Kenney and Christina Synowiec of Results for Development Institute (R4D) for their extensive support for this research and their quality assurance over all formative research deliverables. We would also like to thank Soknea Oun and Sophea Touch for their work in leading in-country focus group discussions and for transcribing and translating transcripts. Next, we would like to thank Drs. Jorge Ugaz of Mathematica Policy Research and Tulika Narayan of Abt Associates for their thought leadership in identifying potential research questions, their quality assurance reviews of formative research documents, their work facilitating theory of change and results framework discussions with Family Care First Cambodia, along with Ms. Kenney and Ms. Synowiec. Next, the team would like to thank Dr. Ugaz and Jeremy Brecher-Haimson of Mathematica for their work in drafting the literature review for this evaluation. We would also like to acknowledge Drs. Stephen Bell and Gael O’Sullivan of Abt Associates for their quality assurance reviews of this work. We would like to thank the Cambodian Children’s Trust (CCT), Friends International (FI), and Save the Children for their support in conducting this research, including their review of the research protocols. Finally, we would like to thank CCT, FI, Save the Children, ReThink Orphanages (RO), the Commune Committee for Women and Children (CCWC), Better Volunteering Better Care, and Intrepid Travel Foundation for their support in identifying sources for this research. Cover photo taken of female Focus Group Discussion participants in Cambodia, with their approval, courtesy of Emily Kenney of R4D. 1 CONTENTS Acronyms............................................................................................................................................................................. 2 Introduction ........................................................................................................................................................................ 3 Development Context ................................................................................................................................................. 3 Project Background ...................................................................................................................................................... 4 Rapid Feedback Monitoring, Evaluation, Research and learning (RF MERL) Background ............................ 4 Evaluation Purpose & Questions .................................................................................................................................... 6 Evaluation Purpose ........................................................................................................................................................ 6 Evaluation Questions .................................................................................................................................................... 6 Evaluation Methods & Limitations ................................................................................................................................. 7 Evaluation Methods ....................................................................................................................................................... 7 Evaluation Limitations ................................................................................................................................................ 10 Findings ............................................................................................................................................................................... 12 Community Behaviors and Perceptions/Attitudes .............................................................................................. 12 Question 1— What types of SBCC campaign messages are likely to resonate with caregivers to change their attitudes and perceptions about RCIs?...................................................................................... 12 Question 2—What communication channels are likely to most effectively reach the greatest number of caregivers? ........................................................................................................................................... 18 Donor and Volunteer Behaviors and Preferences .............................................................................................. 21 Question 1—What types of SBCC campaign messages are likely to resonate with donors and volunteers to keep them from donating to or volunteering with RCIs that engage in poor practices? .................................................................................................................................................................................... 21 Question 2—What communication channels are likely to most effectively reach the greatest number of potential donors and volunteers? ................................................................................................... 25 Conclusions ....................................................................................................................................................................... 30 Community Messages................................................................................................................................................. 30 Community Communication Channels .................................................................................................................. 30 Donor and Volunteer Messages .............................................................................................................................. 31 Donor and Volunteer Channels .............................................................................................................................. 32 Recommendations ........................................................................................................................................................... 33 Communities ................................................................................................................................................................ 33 Donors and Volunteers ............................................................................................................................................. 33 Annexes ............................................................................................................................................................................. 34 Annex I: Data Collection Instruments ................................................................................................................... 34 Annex II: Qualitative Data Codebook ................................................................................................................... 55 Annex III: Bibliography ............................................................................................................................................... 59 Annex IV: RF MERL Overview ................................................................................................................................. 61 2 ACRONYMS CCT Cambodian Children’s Trust CCWC Commune Committee for Women and Children CLA Collaborating, Learning, and Adapting FCF Family Care First FI Friends International HFP Holistic Family Preservation MERL Monitoring, Evaluation, Research, and Learning MERLIN Monitoring, Evaluation, Research, and Learning Innovations MoSVY Ministry of Social Affairs, Veterans, and Youth Rehabilitation MOTIF Mobile Technology for Improved Family Planning RCI Residential Care Institution R4D Results for Development Institute RF MERL Rapid Feedback Monitoring, Evaluation, Research, and Learning RO ReThink Orphanages SBC Social and Behavior Change SBCC Social and Behavior Change Communication TOC Theory of Change TSG Thematic Subgroup UNICEF United Nations Children Fund USAID U.S. Agency for International Development 3 INTRODUCTION DEVELOPMENT CONTEXT Three decades ago, Cambodia saw an increase in the number of orphanages after the rule of Khmer Rouge, which is estimated to have led to the deaths of 1.5 to 3 million Cambodians (Frey 2009). A March 2016 report by the Cambodian Ministry of Social Affairs, Veterans, and Youth Rehabilitation (MoSVY) estimates there are approximately 11,877 children living in 267 orphanages or residential care institutions (RCIs)1 in Cambodia (MoSVY 2016). Not only are these numbers likely “severely underestimated” (Berens et al. 2015), but they continue to increase (MOSVY 2016). Yet, estimates show that only about 23 percent of the children living in RCIs are actually orphans (Fuller 2014). Most come from vulnerable households and are brought there by their parents or extended family to: relieve a financial burden, as a result of their parents’ divorce, or because their parent(s)’ believe the children will receive better care or education from the RCI than they would at home. Also, some RCIs actively advertise for and recruit families to send their children to the institution (MoSVY 2016). Yet, substantial evidence has demonstrated there are significant negative impacts on children living in RCIs. Several studies have found that, compared to family care, living in a RCI can be detrimental to a child’s social, physical, intellectual, and emotional development and can have long-term effects that last well into adulthood (MoSVY 2016). Other studies have shown that education is actually poorer in RCIs, and the risk for physical and sexual abuse is heightened compared to home care (Berens et al. 2015; Guttentag 2009; Browne 2009; Nelson et al. 2007; McGehee and Andereck 2009). The effects are more severe the longer a child remains in the RCI and the younger the child was at admission (Lumos 2015; Smyke et al. 2007; Nelson et al. 2007). RCIs in Cambodia have benefited from both steady demand from vulnerable families who are unaware that RCIs can be damaging to a child and from the growing supply of tourism volunteers (or volun￾tourism)2 and donors (Farley 2015, EMC 2016). UNICEF reports that “Cambodia’s growth in residential care is also attributable to the wealth of support from overseas donors, who with the best intentions, provide support and funding to children in orphanages, unaware of alternative family and community￾based care options.” The organization found that almost all RCIs are funded by expatriates, which incentivizes the RCIs to promote “orphanage tourism” to attract more donors, fueling the problem. UNICEF also found that most Cambodian RCIs accept short-term volunteers, which can negatively 1 UNICEF defines RCIs as “group living arrangements for children without primary caregivers or whose biological parents are unable to care for them. It is meant to provide 24-hour care by paid staff, meeting children’s basic needs of shelter, food, clothing and education. These can be places of safety for emergency care, transit centres in emergency situations, and all other short and long-term residential care facilities including group homes. Other terms used for this form of care are institutional care and orphanages” (UNICEF 2011). 2 Our literature review attempted to define volun-tourism as: “An often-cited definition of tourism volunteers describes them as “those who, for various reasons volunteer in an organized way to undertake holidays that might involve aiding or alleviating the material poverty of some groups in society, the restoration of certain environments or research into aspects of society of the environment” (Wearing 2001). In the particular case of tourism volunteers working in orphanages or RCIs in Cambodia, their motivations are in general well-intentioned — looking for personal growth and self-actualization, wanting to “give back to the world and make a difference,” seeking good personal experiences, absorbing a local culture and, in the case of young people, filling a gap year, usually taken before the college years (Brown 2005; Proyrungroj 2014; Butcher and Smith 2010). It is expected that RCIs will rely more on tourism volunteers in the next few years, as large institutional donors increasingly move away from funding RCIs, preferring to fund child-care options with better practices (BCN 2014).” 4 affect children when the volunteers go home, resulting in developmental issues in some cases (UNICEF 2011). PROJECT BACKGROUND Working with the Royal Government of Cambodia, the United States Agency for International Development in Cambodia (USAID/Cambodia) began funding the $6.5 million Family Care First (FCF) Cambodia Project in September 2015 to help address this problem. The FCF Cambodia project is a multi-stakeholder initiative that includes 27 implementing organizations and seeks to develop a comprehensive care system in Cambodia to prevent family-child separation and promote the re￾integration of children from RCIs to families. Over the next several years, the project will develop and test activities through four key thematic subgroups (TSGs)3: 1) Government system strengthening, 2) direct response to children and families through improved social service workforce, 3) prevention of unnecessary child-family separation, and 4) transformation of the institution-focused care model. This report focuses on experiments related to TSGs 3 and 4. More specifically, as part of the FCF Cambodia Project, the Cambodian Children’s Trust (CCT) and Friends International (FI), two Cambodian nonprofit alternative care institutions, are planning to carry out campaigns to build awareness and influence behavior with regards to family-based care and RCIs. Working under TSG3, CCT’s holistic family preservation (HFP) model is a social and behavior change (SBC) intervention that will include a community social and behavior change communication (SBCC)4 campaign to improve community knowledge of the harms of RCIs and to change the behavior of vulnerable families and local authorities towards RCIs. Under TSG4, FI plans to implement an SBCC campaign to influence the behavior of individuals and organizations that donate time or money to RCIs. In line with its revised collaborating, learning, and adapting (CLA) approach, USAID conceived of FCF Cambodia as an adaptive project, focused on piloting innovative ideas and quickly testing them to determine which are the most effective and efficient before scaling them for impact. Furthermore, according to USAID, FCF Cambodia is the first in a series of potential country-level FCF projects that the agency plans to support under the FCF Initiative. The FCF Initiative uses a "collective impact" model, which brings together stakeholders from different sectors to solve a specific social problem—in this case, it brought together organizations in Cambodia that might help to reduce child-family separation. The model is relatively new, and USAID hopes to build strong evidence of the large-scale impact such a model can have in order to guide future action. RAPID FEEDBACK MONITORING, EVALUATION, RESEARCH AND LEARNING (RF MERL) BACKGROUND USAID signed a cooperative agreement with a consortium consisting of Results for Development (R4D) Institute, Abt Associates, Mathematica Policy Research, and the University of Norte Dame in September 2015 to implement the Rapid Feedback Monitoring, Evaluation, Research, and Learning (RF MERL) Project. RF MERL is an initiative of the Monitoring, Evaluation, Research and Learning Innovations Program (MERLIN) through the U.S. Global Development Lab in partnership with the Bureau for Policy, Planning, and Learning and the Bureau for Global Health. It is an innovative pilot initiative that seeks to 3 TSGs are essentially subgroupings of FCF Cambodia Consortium members who are working together to implement activities within one of the thematic areas. 4 SBC is a term used by USAID to refer to, “Interventions that seek to change behaviors by addressing factors such as knowledge, attitudes, and norms.” 5 apply proven evaluation methods to test the effectiveness of specific components of an activity against alternative intervention options. The approach involves rigorously testing the success of two or more alternative intervention options at achieving short-term outcomes. This is done in rapid cycles, occurring within months rather than years (which is how long a typical impact evaluation takes) to allow for timely feedback and course adjustment earlier than is typically done using standard program evaluation methods. RF MERL is structured to allow USAID to work with partners collaboratively to identify, design, and test cutting-edge solutions to more effectively understand and measure the impacts of development programs before scaling them up. The RF MERL Consortium began working with FCF Cambodia in early 2016 to identify untested, innovative ideas that FCF Cambodia wanted to test before scaling their intervention model. Together, USAID, CCT, FI, and the RF MERL Consortium identified two key areas within the FCF Cambodia theory of change (TOC) for which very little evidence exists but for which there is the potential for significant impact. Those two areas are the community and donor/volunteer behavior change. Behavior change is key to the FCF Cambodia TOC, but little evidence exists for how to achieve that behavior change. It is thought that SBCC campaigns may help to contribute to changes in behavior, but little rigorous evidence exists on the effectiveness of such campaigns. As such, CCT and FI are seeking support from RF MERL to test various SBCC messages and communication channels to identify which result in the most significant changes to proximal outcomes. CCT and FI, in coordination with other FCF Cambodia TSG members, will then scale up the campaigns shown to have the greatest impact, which should ultimately act to increase the efficiency with which FCF Cambodia is able to decrease unnecessary child-family separation. The first step in conducting these tests, or “Feedback Experiments,” is a formative evaluation to identify promising message themes and communication channels for both a community SBCC campaign and a donor/volunteer SBCC campaign. This report is the result of that first step and will be followed by a memo that presents USAID, Save the Children, CCT, and FI with potential Feedback Experiments to determine which campaign options are most effective. Following that memo, CCT and FI will pilot two or more options for each of the campaigns, and RF MERL will test those options using rigorous experimental or quasi-experimental designs to determine which are the most effective. This report is laid out to provide the evaluation purpose and questions first; followed by the evaluation methodology and limitations; and then findings and conclusions by evaluation question, and finally recommendations for the potential SBCC campaigns. 6 EVALUATION PURPOSE & QUESTIONS EVALUATION PURPOSE The purpose of this formative evaluation is to identify possible SBCC campaign messages and channels that might resonate with Cambodian communities and donors/volunteers of Cambodian RCIs and that can also be compared through Feedback Experiments. As such, the key users of the report are USAID, Save the Children, CCT, FI, other TSG partners, and the RF MERL Consortium. EVALUATION QUESTIONS In coordination with the other key stakeholders, the RF MERL Consortium identified two key evaluation question themes—questions about Cambodian community perceptions/attitudes and questions focused on international donor and volunteer perceptions/attitudes regarding RCIs. Both sets of questions follow: Community Behaviors and Preferences 1. What types of SBCC campaign messages are likely to resonate with caregivers to change their attitudes and perceptions about RCIs? 2. What communication channels are likely to most effectively reach the greatest number of caregivers? Donor and Volunteer Behaviors and Preferences 1. What types of SBCC campaign messages are likely to resonate with potential RCI donors and volunteers to change their attitudes and perceptions about RCIs? 2. What communication channels are likely to most effectively reach the greatest number of potential donors and volunteers to RCIs? 7 EVALUATION METHODS & LIMITATIONS EVALUATION METHODS The RF MERL Consortium began the formative evaluation with a desk review of literature that sought to identify the typical characteristics of Cambodian RCI donors and volunteers; the channels of communication by which Cambodian community members and RCI donors and volunteers hear about and come to be involved with RCIs; what research currently exists on community, donor, and volunteer attitudes about and perceptions of RCIs in Cambodia; and what SBCC campaigns currently exist.. The desk review covered both community and donor/volunteer topics. Following the desk review, the RF MERL Consortium collected primary data in the field using the following methods: Community Perception Data Collection and Analysis Methods Instruments—We collected data on Cambodian communities from two different types of sample groups using different methods—focus group discussions (FGDs) with Cambodian caregivers (adults, age 18 and above who care for children) and semi-structured key-informant interviews (KIIs) with village chiefs, Commune Committee for Women and Children (CCWC) members5, and village volunteers. We developed semi-structured FGD and KII guides to capture community perceptions about children, caregiving, RCIs, and channels of communication (Please find all data collection protocols in Annex 1). We then; translated those guides into Khmer; and then had an independent firm back-translate them into English to ensure the accuracy of the translation. Next, we piloted the FGD protocol with two groups of social workers from five different nonprofit organizations that work in in villages in Phnom Penh and Battambang and with a group of self-identified caregivers in Battambang City, and we piloted the KII protocol with eight village leaders in Battambang City. Ultimately, we found that the term RCI was difficult to translate, as described in the findings section of this report. This meant that our FGD and KII facilitators had to ask respondents to explain what type of institution they were describing every time the word for care facility came up—to determine whether they were talking about an RCI or an alternative care institution, such as a CCT or FI. Sample—We selected our community sample using CCT’s HFP model village selection protocol since CCT is the group that will carry out the community intervention. CCT plans to work in three different types of villages—villages with “full services,” some with “partial services,” and some with “no access” to services, as defined below. The reason for this is that CCT wants to test whether its HFP model works in areas with more or less access to health and education services. CCT hypothesizes that whether a family has access to such services is likely a key determinant of whether children are sent to RCIs. As such, CCT identified the following service levels by which it will segment its HFP model outreach and interventions. 5 Every commune council (a governmental entity similar to a city council) in Cambodia has a commune committee for women and children (CCWC), which monitors the situation of children and women within the commune and provides advice to the commune council to budget effectively for the delivery of social services based on the needs of children and their families. The members of the committee vary but should be comprised of at least 40 percent women members. 8  Full-access villages: These villages have a primary school within 1 kilometer of the village and a health center within 4 kilometers of the village.  Partial-access villages: These villages have either good access to a primary school or good access to a health center, meaning there is a primary school within 2 kilometers of the village or a health center within 6 kilometers of the village but not both.  No-access villages: These villages do not have a primary school within 2 kilometers of the village nor a health center within 6 kilometers of the village. Additionally, CCT plans to implement its HFP model in villages in Battambang Province6 that fall within 20 kilometers of their headquarter office. To avoid contaminating CCT’s sample and our possible Feedback Experiment with the questions we planned to ask—including questions about possible messages to discourage child-family separation—we selected our formative evaluation sample from villages that fall outside of CCT’s sample but that otherwise meet CCT’s selection protocol. Since CCT had not yet selected its sample at the time of our research, to select our sample, we drew a circle on a map of Battambang 30 kilometers from CCT’s office.7 We were unable to obtain a map of Battambang’s villages and, thus, used one that included the communes8 instead. As such, after drawing our circle, we eliminated all communes that fell even partially within that 30 kilometer radius. We then created a sample frame that included all of the villages from the Battambang communes that fell in their entirety at least 30 kilometers from the CCT headquarters. Next, we segmented that sample frame using the service access criteria described above—with two “partial-access” segments—one with more access to health services and one with more access to education services. Finally, we randomly selected one village from each of the service areas and three alternates for each service area in case a village chief or village volunteer were not available to meet with our team on the day we visited the village. Our original sample included the villages listed in Table 1. Table 1: Original Community Sample Service Level Village ID # District Commune Village Full-Service 2090503 Samlout Samlout Ou Chrab Partial-Service, More Education 2140101 Rukh Kiri Preaek Chik Seim Partial-Service, More Health 2090506 Samlout Samlout Bueng Run No-Service 2040712 Bavel Khlaeng Meas Bou Sangreach Upon reviewing the sample, our local team determined that two of the villages—Ou Chrab and Bueng Run—were more than 50 kilometers from Battambang, making them likely not very comparable to the villages CCT planned to work in (which would have made our formative research findings less generalizable to areas where CCT does plan to work in the future and where we intend to possibly conduct a Feedback Experiment. As such, the team selected villages from the alternate list that were within 30-50 kilometers of CCT’s headquarters as replacements. Our ultimate village sample follows. 6 Provinces in Cambodia are similar to U.S. states. CCT’s final project activities will all occur within Battambang Province. 7 We chose 30 kilometers–10 kilometers more than CCT’s sampling perimeter—to ensure that we did not inadvertently select a village within CCT’s sample. 8 Communes are the third-level administrative division in Cambodia. They are the subdivisions of districts and consist of between 3 and 30 villages depending on the population. 9 Table 2: Final Community Sample Service Level Village ID # District Commune Village Full-Service 2080305 Sangkae Ta Pon Ta Pon Partial-Service, More Education 2140101 Rukh Kiri Preaek Chik Seim Partial-Service, More Health 2070508 Rotonak Mondol Reaksmei Songha O Khmum No-Service 2040712 Bavel Khlaeng Meas Bou Sangreach Data Collection and Analysis—We conducted at least three KII interviews and two FGDs within each of the villages. We completed KIIs with the village chiefs, village volunteers,9 and CCWC members in each village, except in one village where the village chief was not available;10 as a result, we met with the deputy village chief. In addition to these 12 individuals, we also met with another deputy village chief in one village, which brought the total number of key informants to 13. In addition to the key informants, we selected 4-6 female FGD participants and 4-6 male FGD participants from each village. Village chiefs and village volunteers helped to identify FGD participants and were asked to locate individuals that met the following criteria:  They are parents or caregivers (including grandparents in cases where the parents had migrated)  They are over the age of 18  They have (or regularly care for) at least one child under the age of 18  They have a variety of income levels, community participation levels, family sizes, education levels, etc. Ultimately, we had more than six women show up for most of the FGDs, which meant that we ended up with a sample 21 male caregivers in four FGDs and 34 female caregivers in four separate FGDs. Data Collection and Analysis—Our local team members facilitated, recorded, and took notes for each FGD, and we hired a local firm to transcribe and translate all of the data. Next, the RF MERL team members coded all of the data and analyzed it using NVivo (See Annex 2 for the codebook). Donor and Volunteer Perception Data Collection and Analysis Methods Instruments—We collected data on potential RCI donors and volunteers through two different methods—via an online survey of individuals listed in CCT, FI, ReThink Orphanages (RO), Intrepid Travel Foundation, and Better Volunteering Better Care’s (BVBC’s)11 contact databases and via semi￾structured key informant interviews (KIIs) with individuals identified as “positive deviants” by CCT, FI, 9 Village volunteers are individuals elected in each of the villages to serve as the main point of contact between the village members and the village chief. The village chief works out of the Commune office, but the village volunteer is based out of his/her home in the village, which makes it easier for villagers to communicate with him/her. 10 The village chief wasn’t available because he left town, suspected of stealing money and goods that a nonprofit had given for the community. 11ReThink Orphanages, Intrepid Travel Foundation, and are all Cambodian nonprofit organizations that are members of the FCF Cambodia Consortium. 10 CCFO, M’Lup Russey, Save the Children, and USAID. We defined “positive deviants” as past RCI donors/volunteers who went on to donate to or contact an FCF partner to learn more about the harms of RCIs or the importance of family-based care. We developed the survey and KII guides in English since they were targeted at expatriate donors and volunteers and piloted them with FCF Cambodia TSG partner staff. All donor and volunteer data collection instruments are included in Annex 1. Sample—As shown in Table 3, CCT, FI, RO, Intrepid Travel Foundation, and BVBC sent our survey link out to 1,328 contacts from their listservs. Of those, 94 people clicked on and responded to the survey, but 8 of those only completed the demographic portion of the survey. As such, our total viable sample was 86. None of those individuals who received the email with the survey link via Intrepid Travel Foundation or BVBC completed the survey prior to the survey closing. Our final sample included 58 individuals from CCT’s listserv, 19 from FI’s, and 9 from RO’s. CCT, FI, CCFO, M’Lup Russey, Save the Children, and USAID identified 16 “positive deviants” for the KIIs, and one of the interviewees recommended another “positive deviant” whom we added to the total sample frame through snowball sampling. We contacted all 17 of the positive deviants to request an interview; all responded, and we were able to schedule interviews with all but seven of those for a total sample of 10 positive deviants. We conducted all interviews in English, audio-recorded the interviews, and took verbatim notes. Table 3: Donor Survey Sample Organization Total Surveyed Segmentation Method of Distribution Friends International 787 Donors from the past five years Specific Email Cambodian Children's Trust 397 Current monthly donors who are also subscribed to newsletter Embedded Link with December Newsletter and follow up email Better Care Network 100 Current individuals on the BVBC working group Specific Email ReThink Orphanages 15 Current individual donors to RO Specific Email Intrepid Foundation 29 Donors to Cambodian projects Specific Email Data Collection and Analysis—We analyzed the survey data in Excel, due to the limited sample size. We used Excel formulas and pivot tables to analyze the quantitative results and coded the qualitative results to allow for analysis based on themes. Due to time constraints, we were unable to transcribe all of the positive deviant interviews in their entirety, but we used the available transcriptions to validate verbatim notes and fill in gaps where necessary. We then coded and analyzed the notes in NVivo (See Annex 2 for the codebook). EVALUATION LIMITATIONS Community Perceptions Evaluation Limitations There are several limitations to the community portion of the research, as described below: Selection Bias—We sampled a limited number of villages due to time and resource constraints, which means our total sample is not representative of all possible sample villages. Further, the FGD sample 11 may have been biased since we had to rely on village leaders to select participants; however, this was necessary given time, resource, and cultural constraints. Ideally, the participants would have been a group with all the same characteristics as those individuals who will fall in CCT’s service village. However, it is possible that village leaders named individuals they thought might provide the types of answers we wanted to hear – e.g., RCIs are bad. Next, as described above, we selected villages that fell 30-50 kilometers outside of CCT’s headquarters. While we tried to ensure the villages matched the CCT villages in terms of access to services, there may be other observable or unobservable differences between the villages that make them different from the CCT villages. Response Bias—While there did not appear to be any stigmas associated with discussing RCIs, by discussing a potentially sensitive topic in FGDs rather than through individual KIIs or surveys with community members, there may have been some response bias based on the fact that respondents had to voice viewpoints publicly. Furthermore, we did not purposively select caregivers who had firsthand experience with placing a child in an RCI. While such individuals were included in the final sample, their responses may have been biased by participating in focus groups with others who had little experience with or indirect exposure to RCIs and vice-versa for those who had never placed a child in an RCI. Nonetheless, we used FGDs in an attempt to understand general community perceptions of RCIs. Donor and Volunteer Perceptions Evaluation Limitations We faced similar limitations for our donor and volunteer perception research, as described below: Selection Bias—The primary limitation for the donor/volunteer piece arose from its sampling design. We sent the survey through FCF Cambodia members’ listservs. While this was necessary to feasibly identify respondents with exposure to Cambodian RCIs, the respondents’ exposure to organizations that do not support RCIs may have biased our sample towards individuals with knowledge about the harms of RCIs. Additionally, by using an online survey as our main means of reaching potential donors/volunteers, we likely introduced a bias toward individuals who would recommend online strategies as the best channels to reach potential donors/volunteers rather than in-person strategies such as at Cambodian hotels. Response Bias—While the whole purpose of interviewing positive deviant respondents was to identify people who had donated to or volunteered with RCIs in Cambodia and later changed their donation/volunteer practices, the fact that these donors/volunteers came from FCF contacts and knew about the purpose of FCF Cambodia’s work means they may have felt compelled to provide answers they knew FCF members would want to hear. In an attempt to avoid this, we assured all respondents that their responses were confidential and would not be reported back to FCF Cambodia members. The survey selection bias issue described above (that we sent the survey through FCF Cambodia member listservs) also likely resulted in some response bias in our survey since survey participants may have felt compelled to respond to the survey with what they thought FCF Cambodia members would want to hear. This was likely especially problematic for this group because even though the survey included language assuring respondents that their responses were confidential and would not be shared with FCF Cambodia organizations, participants received the survey link from those groups, which may have made them less likely to believe their answers were truly confidential. 12 FINDINGS COMMUNITY BEHAVIORS AND PERCEPTIONS/ATTITUDES QUESTION 1— WHAT TYPES OF SBCC CAMPAIGN MESSAGES ARE LIKELY TO RESONATE WITH CAREGIVERS TO CHANGE THEIR ATTITUDES AND PERCEPTIONS ABOUT RCIS? Parental Aspirations for Children To understand what parenting decisions concern caregivers most, we asked whether caregivers have hopes for their children and to describe those hopes. Caregivers in FGDs and KIIs predominately reported that they do have aspirations for their children. One male focus group participant said, “We have hope, that’s why we are trying every day.” Most commonly, respondents (6 FGDs, 8 KIIs) reported that they, or parents generally, hope their children will study and receive a good education. They described studying hard and getting an education as prerequisites for getting a good job, which they defined as anything not involving farming, giving examples like being a teacher or working for an NGO. Respondents in six of the FGDs and six KIIs described such occupational aspirations for children. Challenges Families Face Despite respondents’ hope for their children, they described a number of challenges that inhibit their ability to ensure that their children can achieve those hopes or that otherwise negatively affect their children. Specifically, respondents reported problems related to ensuring their children can receive an education, socio-economic issues, and concerns about childcare, drug and alcohol abuse, health, food access, and violence. The most common challenge reported by respondents from both FGDs and KIIs was ensuring their children have access to education (6 FGD, 8 KII). Proximity to a secondary school is often an issue, as described by a female CCWC member, “Primary school is not too far, but if they want to study high school they have to travel a long way to school. They are poor and don’t have enough money to buy a vehicle for traveling. That’s why most of them quit studying when they finish grade 6.” Another major issue that prohibits children from having enough time to study and sometimes results in them dropping out of school is demands on their time to either provide labor within the household or work outside the household to earn money. Several of the other major familial challenges described by participants also affect children’s ability to study and receive an education. Respondents cited lack of money, migration, poor nutrition, limited parental attention and support for education, parental conflict, and exposure to drug or alcohol abuse as factors that inhibit children’s studies. One female CCWC member summarized many of these issues, saying, “Parents do not have any strength to encourage their children. Parents want their children to work at the age of 10, and they want them to stop studying and earn money to support the family. When they migrate to another place, the children follow the mother so the children won’t be able to go to school and study. And, the primary challenge is the family’s economy. The other one [that] is also a challenge is violence in the family… All of these challenges are the entire major problem that effect children in their daily school life.” The second most commonly reported challenge households face with regards to children is socio￾economic problems. Respondents from four FGDs and seven KIIs said that poverty is a major issue for families. To address poverty issues, individuals or families often migrate for work, according to nine key informants and respondents from one FGD, though that is not the only reason people migrate. 13 When parents or caregivers migrate, respondents described four childcare strategies. One is that one parent migrates (usually the husband) and the other stays home and cares for the children (usually the wife). Another is that both parents migrate together and leave the child with his/her grandmother. As one respondent noted, “Some parents leave their children for long because sometimes the parents have to go to work in Thailand; they have to leave children with their grandparents.” Another option is that both parents migrate and take the children with them. “If they don’t have grandparents, they have to bring their children with them because they are afraid others would harm their children; they don’t trust others to take care their children.” A minority of respondents mentioned placing children in an RCI as another childcare strategy parents use when they migrate (For more details, see the section below on placing children in RCIs). Outside of childcare issues due to migration, respondents mentioned childcare in general as the third major challenge facing families. Respondents from two FGDs and six KIIs said families often face issues of ensuring sufficient quantity and/or quality of childcare for children. Respondents noted that, typically, mothers and grandmothers provide most childcare. If the mothers or grandmother is not available, the father usually fills in. Older siblings and other relatives or neighbors also sometimes provide childcare, particularly when parents are absent for short periods. However, childcare can become an issue when families split up due to migration, divorce, or the death of a parent. Several respondents noted that childcare becomes more challenging when children live in single-parent households. For example, a female village volunteer said, “When children just have only their mother, and their mother is also busy with work, those children have nothing to eat and can’t go to school because they don’t have any other [person] to look after them.” Another woman, a CCWC member, indicated a less-tangible impact on children, saying that, “Children do not have a family warmness” when they are left with only their mother. A few caregivers also said that childcare can be challenging because their young children do not listen to them. For example, one female focus group respondent said she “Advises children, communicates to comfort the children, [but] children don’t listen to [her].” Other challenges respondents identified included drug and alcohol abuse, violence, and child health issues. Respondents from three FGDs and three KIIs said that families face issues of either parental or child drug or alcohol abuse, both of which can negatively affect children. A similar number of respondents said families face issues of violence such as child abuse, sexual violence towards children, and violence in general. Four key informants and participants of four FGDs reported that child health issues can be challenging for families. In some cases, parents and caregivers said that children’s illnesses are a problem, and in others, they said they have trouble paying for healthcare when their children are sick. Relatedly, some respondents (2 FGDs, 3 KIIs) said that access to water, including clean water, is a problem as well as access to food (1 FGD, 6 KIIs). Community Services Available When describing services families have available to help address the challenges they face, respondents most commonly reported they had access to health services, services from government and nonprofit organizations, and education. Respondents from eight FGDs and five KIIs said health services, such as health clinics or village volunteers, were available in their community. Respondents from two FGDs and five KIIs said that the government provides support, such as commune funds for healthcare and financial support to poor households in their community. Four FGD respondents and five key informants reported that local and/or international nonprofit organizations, such as CARE and World Vision, provide services for women and children (e.g., legal support for victims of violence and sexual abuse, advocacy on women’s and children’s rights, material support for children—like books, food aid after disasters, microloans, vocational training, and/or support for school fees), among other supports in their communities. Only respondents in one FGD and two KIIs mentioned schools or kindergartens as 14 services available in their community that help families with challenges. Notably, when asked about the services available in their communities to help families deal with challenges, respondents did not list RCIs. Naming RCIs In planning for a possible SBCC campaign targeted at communities, it is important to note the exact language respondents use when they talk about aspirations, challenges, services, and especially their perceptions of RCIs. One challenge the research team faced while in the field was differentiating between when a respondent was talking about an RCI versus an alternative care institution like CCT or FI. For instance, a few respondents at first discussed kindergartens, other non-residential childcare or educational centers, or health centers when asked about whether there was an RCI in their community. As such, we have recorded the types of words that respondents used to describe RCIs. We found that, typically, respondents refer to RCIs as “children care centers,” “childcare centers,” or “centers.” However, they also used terms like “organization,” “orphanage,” and “orphan center.” 12 Knowledge and perceptions of/attitudes about RCIs The majority of respondents who reported on RCIs did so based on secondhand knowledge. Only two FGD participants reported having placed children in RCIs themselves. Participants inconsistently reported whether or not there was an RCI in their community. This inconsistency may be a result of varying knowledge about RCIs or what distance they were including as part of their communities or even an issue of how people were defining RCIs. The majority of participants (19 respondents) from four of the communities we visited said that they do not have RCIs in their community. Four respondents across three communities reported having an RCI in their village. It was not possible for the research team to confirm whether there were RCIs in any of the communities, given the difficulties in identifying unregistered RCIs. Most respondents had heard of RCIs and knew something about them. Respondents typically described RCIs as a place where children live and are provided with education, care, and food. People also described them as a place where orphans can receive care, but not as a place exclusively for orphans. Respondents also did not necessarily see RCIs as permanent care facilities or facilities that require ties to be permanently severed between children and families. A few people said that they had heard about or seen either parents visit children at RCIs or children visit home (e.g., for holidays). A few caregivers also said that a child might go to an RCI for just a few years, but the child can also come home if they do not like being at the RCI.13 In 4 FGDs and 8 KIIs (4 village chiefs/deputy chiefs, 3 village volunteers, and 1 CCWC member) respondents said they have positive impressions of RCIs and mentioned few harms for children or families resulting from RCI care. A few respondents said that parents even pay to have their children attend an RCI. They cited job (3 FGDs, 5 KIIs) and educational opportunities (3 FGDs, 5 KIIs) as the primary benefits of RCIs. Women raised both job and educational benefits of RCIs, more frequently than men, although women in general gave more detailed responses about why RCIs are beneficial compared to men. For job opportunities, respondents said that the benefit is children’s improved ability to get a job after being in an RCI because of the education they receive, from vocational training, or from direct support from the RCI to find a job. Respondents described two main ways that RCIs provide better educational opportunities and outcomes for children. One is that they provide an option for secondary school education, which children often do not have near their homes. So that makes 12 In Khmer, these terms are: children care centers មណ្ឌ លថែទ ាំកុមារ, childcare centres មណ្ឌ លថែទ ាំកុមារ, centers មណ្ឌ ល, organizations អង្គការ, orphanage មណ្ឌ លកុមារកាំព្រា, and orphan centers មណ្ឌ លកុមារកាំព្រា. 13 Research shows that this is true in some cases—parents do not have to sever ties. But, some RCIs do require this. 15 some families “want to enroll [their children] in the center to continue to study,” according to one female CCWC member. The second reason often given that RCIs help children to receive a better education is that children study more at RCIs. Caregivers attribute that to both the increased level of attention children receive at RCIs and the strict discipline RCIs provide. One male focus group participant said that at home, children “Can do whatever they want based on their desire; parents can’t force them.” Another male focus group participant pointed to parents’ circumscribed agency in influencing children and how it relates to educational attainment, although not in specific reference to RCIs. He said: “I would like my children to follow their dreams and hopes, because some child has a natural skill [from] their birth then we have to follow them but [cannot] force them. Even if parents want their children to have a bright future, if we force them, then it would be a disaster in the end. The children in the village, they can only farm, and they do not get a good education. And, [only] approximately 20 percent of children finish grade 12.” One male village chief explained the difference between being at home and in an RCI, saying, “They are different because the parents have no free time to observe their children, while the children at the organization are looked after properly and restrictedly. So that the children at organizations could study regularly.” One female focus group participant also said that in RCIs, “The child will grow up in the environment that makes them want to study,” rather than forcing them, and “Therefore, when they grow up, they will be educated and won’t get involved in walking in the wrong path,” (referring to when children are involved in negative things, like crimes or drugs). Literature suggests a third possible reason why RCIs may be perceived as able to provide better access to education than households, and that is due to the cost, as described below under the section on Placing Children in RCIs. Some also cited other benefits of RCIs, such as providing childcare (2 FGDs), a place to stay (2 FGDs), and food for children (3 FGD, 6 KII). As with the other benefits discussed above, women more commonly specified these benefits compared to men. A minority of respondents explicitly said that RCIs care for children better than families. As one female focus group participant explained, “I think for the children who live in the RCIs, their life is better because they get so much attention, such as their food, their clothes, and they are really full of love from the RCIs.” For example, a male village chief said, “Honestly, at the orphans center they take care of children better than at home.” A male focus group participant compared family and RCI care and related the better care at RCIs to living conditions and the quality of caregivers. He said: “It’s different. At the center, they take care of children better than us; [for instance], clothes are better. We live in the village, and it’s dirty. They live at the center. It’s cleaned, and they don’t use violence on children. It’s because the center selects good children-sitters to take care of the children.” Some respondents reported negative impressions about RCIs or downsides to RCI care. While only participants from one male focus group raised negative impressions or effects, seven key informants did (5 female, 3 male). Most commonly, respondents said that children will be sad, miss their parents, have less time with their family, or their parents will miss their children if they are at an RCI (1 FGD, 6 KII). One male focus group participant said that when he asked a child he knew “about staying at the center, [the child] cried because he was so mad at his mother,” who took him to the RCI. These respondents did not always think that these negatives were severe or outweighed the benefits of RCIs. One male village chief said, “The disadvantage is that they have less time with their family, but they get a lot of benefit, such a better education, which they could [use to] get a good job in the future.” The other negative aspects reported are that RCIs may not care for children well if there are too many children 16 living there (1 KII) and that RCIs can be too strict (1 KII). Aside from these negatives of RCIs, respondents provided little other indication that they associated family care with improved overall care or emotional or psychological well-being or happiness. Respondents did not express awareness about the potential for exploitation of children in RCIs or the fact that some RCIs provide very poor living conditions. In a few instances, respondents described situations surrounding children’s placement in RCIs that sounded as though they could have been exploitative; yet, the respondents did not recognize that. In one case, the respondent said that a mother decided to send her 8-year-old daughter to an RCI after the little girl was raped and the adult perpetrator threatened to kill her if she told anyone. The respondent said: “We could not afford the money to sue the guy who was threatening the young girl until an organization in Phnom Penh came to take the young girl to study skills such as sewing and massage for a job in Phnom Penh. Because this village is not safe, then her mother decided to let her daughter go with the organization.” UNICEF also found that vulnerable families and communities in Cambodia are quite aware of the benefits of family-based care arrangements for children, but there is contradictory information regarding how much they know about the risks of institutionalized care (UNICEF 2011, Jordanwood 2016). A few respondents noted restrictions on who RCIs will take, saying that children often have to be poor, girls, or below a certain age (e.g., younger than 10-12 years old). Placing children in RCIs When asked about why children were placed in RCIs, respondents replied without hesitation, suggesting there is little to no stigma associated with placing children in an RCI. In fact, as described above, while most of the responses are based on respondent perceptions or knowledge of other households that sent children to RCIs, two FGD respondents reported from firsthand knowledge, having placed a child or grandchild in an RCI themselves. Directly in line with challenges identified by respondents, FGD participants and key informants across all four communities primarily said that children are placed in RCIs because the facilities provide better education for children (4 FGDs, 6 KIIs) and because families are too poor to care for children (4 FGDs, 5 KIIs). A few respondents indicated that some people may place children in RCIs just for the educational opportunity, even if their family has sufficient resources. For example, one female focus group participant said: “Some families are not really poor, but they wish to educate their children the way the orphanage does because when the children are home, they don’t study hard. When they are with parents, they don’t listen to their parents at all, but when placed in orphanage, there are strict rules and principles.” These findings are further supported by evidence from the literature review, which found that families usually report sending their children to an RCI due to a combination of factors, most of them related to the family’s lack of capacity to provide for children’s education or to poverty issues within the household (UNICEF 2011). The literature also suggests that a child’s education needs are a much better predictor than poverty status of the likelihood that a family will send a child to an RCI. Further, in a survey of more than 1,000 Cambodians, UNICEF found that even though 61 percent of parents surveyed said that families are better at raising children than RCIs, 89 percent of families said they believe children receive a better education in RCIs than they do at home. And, 92 percent of households said that families should send their children to RCIs if they cannot afford school fees and expenses. Finally, UNICEF reported that many households cannot afford schools fees (UNICEF 2011). They wrote: 17 “In principle, public education in Cambodia is free. In reality, however, teachers are paid very little and as a result take “informal school fees” from parents. In addition, parents must pay for school books and uniforms. Many also need to provide bicycles to help children get to school in the higher grades when schools are further away. A World Bank report in 2005 noted, ‘The average schooling expenditure per primary school child accounted for 26.5 per cent of non-food spending among poorest households’ (p.64)” (UNICEF 2011). The lack of ability for households to pay school fees, resulting in children being sent to RCIs, is also an issue addressed by a World Bank report. The report shows that costs and fees associated with attending school increase with the age of a child. Total expenses increase from around 117,000 KHR per child to around 215,000 KHR between the first and sixth grade (World Bank 2005), which may help explain why the average age at which a child enters an RCI in Cambodia is 11 years old (Farley 2015), the age when they are finishing basic elementary school. One village chief validated this hypothesis, saying that there are about 10 children from his village who are in RCIs. He said, “They take their children to [the RCI] after they finish grade 6… They have parents, but they are poor; so, they decide to bring their children to the organization.” Of the four villages we visited, only the “full-service” village had a junior secondary school in the heart of the community, but even that community was 10 kilometers from the nearest senior secondary school. The other three communities were at least 7 miles from the nearest junior or senior secondary school and as many as 27 miles away, suggesting secondary school access may be a bigger issue than primary school access. While a few respondents said that families would prefer to stay together, these respondents acknowledged that because of economic hardship or other problems families cannot or should not always stay together. For instance, in trying to describe why it may be difficult to discourage families from placing children in RCIs, one male village chief said, “I think no one wants to separate from each other. They want to stay together, but the living standard is low. If they don’t come to find a job, it’ll become lower and lower, so have only migration only for them.” Similarly, in talking about RCIs, a female FGD participant said, “I think it is the best choice for very poor families. However, those families in the middle class can afford to take care and raise up their children independently; so, they might hesitate to send their children away.” Other factors that respondents said contribute to why children are put in RCIs are inadequate availability of caregivers (6 KIIs), food access (2 FGDs, 6 KIIs), and violence (1 FGD, 4 KIIs). Literature also supports these conclusions, with reports showing that families reported these additional reasons for sending children to an RCI: parents getting a divorce, parents migrating for work, a lack of close relatives in the community, substance abuse at home, and no access to land. Literature also finds that another important indicator that increased the likelihood of a child being placed in an RCI was if that child’s family knew of other families with children at that particular RCI (UNICEF 2011). When asked whether they would be willing to place children in an RCI and why or why not, a minority of respondents discussed reasons why they or someone they know would not want to place children in RCIs. They cited the following reasons (in no particular order): because they want their children to stay home to provide labor for the household (and earn money), to go to school, because they have parents, the RCI is far away, and/or to continue to live with the family. One respondent said, "It is not good to live separately from children.” Another focus group participant said that someone tried to come to take her child, but she did not let her go. She said that she met an Australian who recruits children in her area. She said, “I met her, she came to take my child, but I didn’t let my child go because my child was still small. At that time, my child was only 5-6 years old.” A male focus group participant said, “I don’t let my children go because my children have parents, but about the orphaned children maybe they can go.” Alternatively, another male focus group respondent said, “If the RCIs are far away from my home, I 18 don’t let my children go.” Another FGD participant explained that she thought her nephew should go to an orphanage, but his other aunts and uncles did not let him, saying: “The aunts and uncles of the child do not allow placing him in the orphanage because they said that they pity him though he’s poor in his study, they would raise him up by themselves together. Which is contradictory to my purpose that I want to send him to the orphanage in order to get access to education since every expense is guaranteed by the center. As a matter of fact, we don’t have abilities to support our [own] children; yet, how are we supposed to support that child?” One male village chief said that he sees perceptions beginning to change about RCIs, with more people not wanting to send their children to RCIs than before. He said, “In my village…parents think that it is a new generation. They can keep their children with their parents, and then they can go anywhere.” Others did not point to changing perceptions. Messages to encourage families to stay together We asked both FGD participants and key informants what messages they thought would be effective at encouraging families to stay together—messages about keeping families together, the negatives when families split, education opportunities in communities, or the negatives of RCIs. When asked their opinion, most often respondents across all four communities said the messages they thought would resonate with families were messages about the benefits of families staying together, citing the importance of familial relationships, (4 FGDs, 10 KIIs) and messages on educational opportunities in the community (5 FGDs, 3 KIIs). The respondents who said that messages about educational opportunities would resonate well represented all four communities in the sample—including those from “no-service” and “full-service” areas—suggesting a lack of correlation between distance to primary school and the quality of education available in a community. These respondents included both individuals who said children should be placed in RCIs to improve their educational prospects (3 FDGs, 2 KIIs), and individuals who did not say that (2 FGDs, 1 KII). Four key informants thought that messages on the negatives of when families split up would be most effective. No respondents said that messages focused on the negatives of RCI care would be effective. Despite these results, one female CCWC member explained a dilemma she saw in communicating messages about keeping families together. She said, “Because if their family are really poor, they have no decision. So, they have to send their children to the center. It doesn’t mean they don’t love their children, but they want their children to continue studying. I know how they feel. Even if we explain to them our message, they have no decision. If we talk about happiness in family, it might just make them sad because they want to live together with children but they also want their children to continue studying.” QUESTION 2—WHAT COMMUNICATION CHANNELS ARE LIKELY TO MOST EFFECTIVELY REACH THE GREATEST NUMBER OF CAREGIVERS? Types of media use and access Before looking at preferred channels of communication, it is important to examine what channels are available to most community members. Overall, mobile phones, television, and the radio were the most common media respondents said they use to get information and that people in their communities have access to. They consistently cited mobile phones (4 FDGs, 12 KIIs) and television (7 FDGs, 10 KIIs) as the most widely used media. While mobile phone use is common, respondents noted that not everyone has access to smart phones. They said they use mobile phones for many different purposes, including to make calls, play games, and use Facebook. The literature shows that developing countries often benefit from phone-based interventions because they draw from existing low cost technology while improving 19 the coverage and quality of interventions (Cole-Lewis and Kershaw 2010, Higgs et al. 2014). Cambodia has one of the largest phone penetration rates in the developing world, with 94 percent of Cambodians owning their own phone and about 40 percent owning a smart phone (Phong and Sola 2015). Respondents also reported that they watch television in the evenings, including the news and Thai movies. More than half of FGD participants and about 46 percent of key informants mentioned listening to the radio (5 FGDs, 6 KIIs); however, some respondents said that radio use was not common, and several key informants reported that fewer community members have access to the radio than have access to mobile phones and televisions. One key informant and participants from one focus group said that only older generations listen to the radio. Those who reported listening to the radio said they listen to programs such as the news and Buddhist programs. Some participants also listed the internet (1 FGD, 8 KIIs) and Facebook (2 FGDs, 7 KIIs) as common media available in the community. However, not all respondents who said they use Facebook reported having access to the internet. Children use Facebook for play, relaxation, and news, to a lesser extent. Five key informants said primarily young people use Facebook, rather than adults. One key informant also noted that Facebook can have fake news on it and does not always have accurate information. Sources of trusted information The majority of focus group respondents reported that when they are looking for information they trust, their preferred source is the village chief or other leader (e.g., deputy village chief) (6 FGDs – 3 male and 3 female). Four key informants also listed village chiefs or other leaders as an important sources of information. According to respondents, the village chief primarily shares information via community meetings. The majority of key informants (10 KIIs) but respondents from only two FGDs reported television as a main source of important information. A minority of respondents reported a range of other sources of important information, including radio (1 FGD, 3 KIIs), village volunteers (1 FGD, 2 KIIs), health centers (3 KIIs), other government sources (e.g., communes) (2 FDGs), Facebook (1 FGD, 1 KII), mobile phones (2 KIIs), community members (1 FGD), the internet (1 KII), social media (1 KII), and a community announcement or broadcast (1 KII). Sources of information on support for children When asked more specifically where they get information about services and support for children, male and female respondents identified health service workers (e.g., health center, health volunteers) (5 FGDs, 4 KIIs) and village chiefs (4 FGDs, 1 KIIs) as key sources of information. Some respondents also listed television (1 FGD, 6 KIIs), village volunteers (1 FGD, 5 KIIs), radio (1 FGD, 3 KIIs), nonprofit organizations (2 FGDs, 3 KIIs), communes (1 KII), CCWC (1 KII), teachers (1 KII), and/or other community members (1 KII). Decision-makers who place children in RCIs Respondents said that parents, relatives (1 FGD, 5 KIIs), and village chiefs (7 KIIs) are primarily involved in deciding to place a child in an RCI. Of the key informants who reported that village chiefs are involved in this decision, 2 were CCWC members, 2 were village chiefs, and 3 were village volunteers. Village chiefs may help families to register a child or connect the family with an RCI. In extreme cases, the village chief actively helps facilitate a number of children going into residential care. For instance, in one community, a female CCWC member said, “Because every year, the center needs twenty to thirty children to enroll. They request to the village chief, and the village chief has to ask every family in the village to find children to enroll.” In another example, a female CCWC member said that encouragement to use RCIs can come from higher government officials, saying, “The district governor [may] tell the mother of the children to take the children to the center. Then she goes to work or earn some money, and if one day she wants to take her children back, she can.” CWCC was another group 20 that was cited as sometimes taking children to RCIs (2 KIIs). A few respondents mentioned that people from outside the village sometimes solicit children to take to their RCI (both foreigners and Cambodians). These latter reports are supported by the literature, which suggests that some RCIs actively market for children to be sent to them. UNICEF found that some families reported receiving money and food in exchange for sending their child to an RCI. They also found reports of local government officials actively persuading parents to send their children to RCIs. And, in interviewing staff from RCIs, they found that staff acknowledged having visited communities and – after requesting to be taken to the most vulnerable and poorest families – having targeted children to take to the RCI with the approval of the village chiefs (UNICEF 2011). Channels for messages on keeping families together When asked what they thought are the best methods is to share information on the importance of keeping families together in their communities, respondents by and large cited in-person strategies. Across respondent groups, men and women, old and young, commonly said that community meetings (5 FGDs, 10 KIIs) were the best way to share and receive information and that village chiefs were the best ones to distribute information in these meetings (4 FGDs, 7 KIIs). A few respondents suggested that it might be useful to show advertisements in these meetings to share news (3 KIIs). One noted that the meetings should be 10-15 minutes long so participants do not get bored, and another said that organizers would need to offer gifts to get people to attend. However, as discussed above, respondents also said that village chiefs often help make the decision to place a child in an RCI (7 KII), or refer families to RCIs (1 FGD, 4 KII). Yet, UNICEF found that some village chiefs may be interested in playing a more active role in improving knowledge and influencing behavior of their constituents regarding RCIs (UNICEF 2011). A few respondents mentioned a few other ways to get messages out in communities, including through television (2 FGDs, 3 KIIs), mobile phones (3 FGDs, 1 KIIs), door-to-door campaigns led by village volunteers (3 KIIs), announcements broadcast on the street via microphone (2 FGDs, 1 KIIs), radio (1 FGD, 2 KIIs), health centers (2 KIIs), nonprofit organizations (1 KII), the village commune (1 KII), information boards (1 KII), and Facebook (1 FGD). Far fewer respondents reported that electronic methods would be a good means by which to share information in the community than said that they regularly use such media (see section above). One female CCWC member explained that this is because, “If we inform by television, [community members] won’t understand; so, we have to tell them directly.” Also, while phone penetration rates are high in Cambodia, recent studies suggest mixed results from SBCC mHealth interventions (one of the most common uses of mobile technology in international development). One study in particular reviewed 21 randomized controlled trials and found that more than half of them failed to document significant health improvements from mobile SBCC campaigns (Gurman et al. 2012). Further, in Cambodia previous attempts at using SMS-based mHealth interventions have anecdotally had mixed success. Some have had limited success due to the lack of techno-literacy, general illiteracy, and phones not supporting Khmer script (Marie Stopes International 2014). Nonetheless, some have found initial anecdotal success. For instance, the Cambodian Mobile Technology for Improved Family Planning (MOTIF) program implemented a voice message-based campaign as opposed to a text message campaign after finding that their target population was more responsive to voice messaging in promoting contraception use in Phnom Penh, Siem Reap, and Battambang (Smith et al. 2016). While the RCT being conducted to test the effects of MOTIF does not appear to be complete yet, early indications are that clients appreciate the service and find the support received helpful. Another phone-based Cambodian SBCC campaign focused on diabetes control found success in 21 leveraging existing community-based peer educator networks to efficiently implement its text-based intervention (van Olmen et al. 2013). UNICEF has sponsored studies to explore the best dissemination methods for community SBCC campaigns in Cambodia. UNICEF’s studies found that (i) village chiefs and commune leaders should be targeted as advocates due to the influence they hold in their communities, (ii) community events are effective at attracting large groups from the target population to spread awareness, and (iii) teachers can be effective communication channels on issues related to child protection. The studies also said that media campaigns in Cambodia should target TV, as it has the widest reach, and can be complemented by radio, which tends to reach more females and rural populations (Indochina Research Ltd 2016; UNICEF 2016). Household decision making We asked respondents about household decision making, because understanding the relative decision￾making power of different household members can indicate to whom behavior change messages should be targeted. Responses were mixed. Women from three FGDs and men from three other FGDs said that in their households, husbands are the ones who primarily make decisions. For instance, a male focus group respondent said, “I am the head of the family. I am a husband. I am a father. We discuss in the family and ask the father before making a decision. The mother does not make a call, only me.” One woman said, “For me, I always listen to my husband. [When] I make a decision, it depends on him because he is a leader; so, he is always true, and I always agree with him because I can’t lead.” Yet, women from three other FGDs and men from two reported that in their households, women primarily make the decisions, especially about childcare. For instance, one woman said, “Mothers are the ones who make decisions about childcare because the fathers go out to generate daily income for the family. Mothers are the ones who take care of the children.” One male respondent said that the primary decision maker is his wife because she stays home. He said, “I was a decision maker, but I don’t stay at home; so, I will give to my wife because she stays at home, she can make decisions. If I stay at home, I will be the head of the family.” Alternatively, in six focus groups and one key informant interview, men and women explained that husbands and wives discuss household matters together in their household. Despite all of this, FGD participants did not say that either parent makes the decision to send a child to an RCI. Village chiefs and government officials were the only specific individuals identified as making these decisions. DONOR AND VOLUNTEER BEHAVIORS AND PREFERENCES QUESTION 1—WHAT TYPES OF SBCC CAMPAIGN MESSAGES ARE LIKELY TO RESONATE WITH DONORS AND VOLUNTEERS TO KEEP THEM FROM DONATING TO OR VOLUNTEERING WITH RCIS THAT ENGAGE IN POOR PRACTICES? Motivations for Donating or Volunteering with Nonprofit Organizations The donors and volunteers surveyed reported various motivations for donating to and/or volunteering with CCT, FI, or RO. A total of 90 percent of respondents (77 of 86 people) reported having donated to CCT, FI, or RO. When asked through an open-ended question, the most common reason they gave for donating to the organizations was because they believe in the work (32 percent of respondents). Other reasons for donating are shown in Figure 1 below. 22 Figure 1: Primary Reasons for Donating to Cambodian FCF Institutions Three out of ten positive deviant respondents said they donated and/or volunteered with the FCF Cambodia organization with which they are associated because they wanted to help and do good. Two out of ten respondents had traveled in Cambodia or Asia and felt affinity toward the country. Two out of ten respondents were affiliated with organizations that support RCI alternatives and donated or volunteered because of that involvement. Two out of ten respondents cited the good that the organization was doing and wanting to support it/visit as their motivation. Perceptions of RCIs Positive deviant interview respondents presented a range of perceptions and knowledge of RCIs. They typically reported that compared to when they first learned of RCIs, they now have a more negative impression of them. However, there was a wide spectrum of the level of negativity of impressions. Some respondents had quite negative perceptions, grounded in the idea that RCIs are always harmful to children (whether or not the children are orphans), no matter how well the RCIs are run or how well they provide for children. While three out of the ten respondents (30 percent)—including two former volunteers and one individual who works for a nonprofit organization—reported thinking that RCIs could be okay if they were well run and well-intentioned, they usually indicated that for this to be the case, the RCIs would have to have some way of reconnecting children with their families. In describing how they changed their minds about RCIs, several positive deviant interviewees described a period when they questioned whether they should be donating to or volunteering with a particular organization due to a suspicion that something was not right. During that period, they tried to learn more or confirm whether the organization was good or not. A few respondents found it difficult to determine definitively whether the organization was well-run and well-intentioned. These respondents would have liked to have had more resources available to determine the quality of the organization. A number of the interviewees (5 out of 10) had direct personal experiences with an organization that prompted them to question or change their mind about whether to continue supporting the organization—causing them to look for alternatives to donate to such as FCF Cambodia Consortium members. For instance, a few people volunteered at RCIs and felt that there was something off, that did Believe in the organizations' work 32% [CATEGORY NAME] [PERCENTAGE] Personal experience with organization staff member 20% [CATEGORY NAME] [PERCENTAGE] Visited the organization in Cambodia and liked it 10% Organization was not affiliated with a religion 6% 23 not feel right, or that raised red flags for them (e.g., being left alone with children, children’s strong attachment, no police checks). Some people did not fully sort through how those experiences changed their impression of RCIs and donation/volunteer practices until later on (e.g., after learning about the issue through news coverage of CCT). Another donor started asking questions when he couldn’t find more information about the organization he was supporting online. Donating to or volunteering with an RCI As described in the methodology section of this report, survey respondents were likely predisposed to having negative views of RCIs given that their contact information came from FCF Cambodia member listservs. Nonetheless, of survey respondents who answered the question, 63 percent (49 respondents) said they would consider donating to an orphanage, while 56 percent (45 respondents) would consider donating to an orphanage in Cambodia. Of those who would donate, 35 percent of respondents (11 out of 32) said the top reason why is if they agree with the organization’s model and/or trust the organization. For example, one respondent said, “If that is the best/only alternative for some children, I'd consider it.” Another respondent said “I am a Cambodian and even I don't have much money to help... but I am willing to share what I can (even not regular) to support the orphanage or residential care in Cambodia because it can help them to stay warm and get a chance to study for their future.” Of those who said they would not donate, 27 percent (9 of 33) cited the information CCT, FI, or RO provided against doing so as the top reason they would not consider donating. They were not able to cite the exact messages/campaigns that made them think they should not donate (only the channel, as described below). However, they did give various reasons for not wanting to donate. For instance, some cited concerns about the RCIs themselves, saying:  “Exploitation is rife—are these children really orphans or pawns for profit?”  “Orphanages in Cambodia seem to take children who actually do have families who are financially unable to support them. [I am] unsure if they exist for the benefit of children or organizers.”  “It's an unsustainable and unethical model.”  “Orphanages create orphans.”  “I have learned via CCT that these places often don’t house actual orphans and are used by families as a way of gaining income while child is stuck there.” Others cited the importance of family, saying:  “I believe it is an 'industry.' Children should be supported to stay with their family.”  “I like the idea of helping families to stay together. That's why I donate.”  “Family based care is best for children, as instituted by CCT.”  “I believe that every child has a right to family and therefore donate towards projects which keep families together or implement family-based alternative care.” Finally, others had concerns for the children’s long-term well-being, saying such things as they would not donate “Because of research into long-term problems for children raised in orphanages.” About one-third of survey respondents (38 percent or 29 respondents) said they would consider volunteering with an orphanage. The top reason, spontaneously reported by 6 respondents (21 percent of the 29), for why they would consider volunteering is a desire to contribute time or skills. For instance, one person said, “Again, I would have to really agree with the reasons it existed. Depends on the country, why the kids are there, if they have living relatives, and if those relatives have safe homes.” Others said they would consider volunteering:  “To pass on relevant skills, which would help the individuals attain a better future for themselves, their families, and the country.” 24  “Only if I was convinced that my participation was truly needed, i.e. not for my own sake.”  “[If the RCI] “Seemed to be a safe place, last resort, with good outcomes, and stable homes.” Most respondents (64 percent) said they would not consider volunteering with an RCI. The reason the majority (42 percent) cited for why they would not volunteer was due to concerns about the orphanages and their practices. Nine percent also reported that they wouldn’t volunteer with an orphanage or RCI because they do not believe doing so is beneficial for children. They said children deserve professional, long-term care, not care from short-term volunteers. One respondent said, “Volun-tourism is not healthy or productive for the people and culture of a country. And orphanages are a by-product of a time that we are now well-researched beyond and alternate avenues of care, health and wellbeing are safer and holistically better than any form of child institutionalization.” The literature corroborates the belief voiced my some survey respondents above that most donors probably do not know that Cambodian orphanages do not only contain orphans. One study found that 75 percent of potential volunteers think that all children in RCIs are orphans (Farley, 2015); and even though most of the potential volunteers acknowledge that RCIs may be not the best option overall, half of them are not aware that RCIs can be seriously detrimental to the long-term development of a child, or that they are run as for-profit businesses that must rely on the constant flow of children for their own survival (Little 2015.) This must be at the core of any SBCC effort to prevent child-family separation. While 32 respondents (47 percent) would recommend that friends or family consider donating to or volunteering with a Cambodian orphanage or residential care institution, upon further examination, it turned out that several of them (5 respondents) were confusing RCIs with alternative care organizations like CCT, FI, and ReThink Orphanages. Others said they would only recommend that friends volunteer if it was with an organization they trusted or that had been vetted. The top reason, applying to 53 percent of respondents, that they would not recommend that friends consider donating to or volunteering with RCIs was that they had concerns about the orphanages and their practices. Messages to discourage support for RCIs Four of the ten positive deviants spontaneously reported that they thought that donors and volunteers are generally well-meaning and need to be educated about the “harms of RCIs.” For example, one respondent said: “I think it’s really about education. I think people generally want to do the right thing, and they think they’re doing the right thing when going to an orphanage without thinking about what they’re actually doing because they just haven’t been educated. The more education that can happen, the better.” Three additional respondents alluded to similar ideas, suggesting that messaging should aim to explain the context of and logic against supporting RCIs. For instance, one respondent said, “I think it’s understanding that you can create more orphans, you know, by doing this… [if] you are traveling and supporting organizations that aren’t doing the right thing then you’re just perpetuating it.” One respondent suggested sharing the percentage of children who are not orphans in Cambodian RCIs as a telling data point. Another respondent said that “messaging simply that it’s not good to support an orphanage won’t get anywhere”—there needs to be enough context and content about the situation to convince donors that not all donations are a good thing. Three positive deviant survey respondents also spontaneously suggested that messages should include positive alternative actions for people to take, because they said many people want to help and might be more receptive to learning about positive alternatives rather than just being told to do nothing. One respondent said that the challenge now is that it is difficult to know what these positive alternatives are, 25 saying, “there are a lot of organizations and if you aren’t in Cambodia you don’t know if they’re okay or not okay.” Another said that once donors “understand what they’re funding and [that] they’re drawing children into care, they’re readily open to looking at other options and willing to talk about alternatives for children.” Seven of the ten positive deviants said messages should talk about how RCIs are not safe for children. For example, four respondents said that messages should cite the social and psychological harm that separation from families causes children. Three said messages should emphasize the revolving door of volunteers and the importance of long-term emotional support. They said that educating potential volunteers and donors about the effects of residential care on child development would be helpful. We did not include a question about potential messages for survey respondents because we did not want to contaminate our potential sample for the feedback experiment. Nonetheless, a few themes did emerge amongst survey respondents. Across all questions, a total of 41 survey respondents mentioned the importance of keeping families together, 7 mentioned the importance of keeping children safe, and 7 mentioned concerns over exploitation among children. QUESTION 2—WHAT COMMUNICATION CHANNELS ARE LIKELY TO MOST EFFECTIVELY REACH THE GREATEST NUMBER OF POTENTIAL DONORS AND VOLUNTEERS? Channels for Learning about Donating or Volunteering Messages to not donate/volunteer are most likely to be successful if they go through channels by which people learn about RCIs. When asked how they learned about CCT, FI, or RO, donors and volunteers overwhelming identified television and personal connections. As shown in Figure 2, 38 percent of survey respondents (33 of 86 respondents). reported they learned about CCT, FI, or RO through a TV program; 28 percent (24 of 86 respondents) reported that a friend told them about that organization; and 7 percent (6 out of 86 respondents) reported they learned about that organization through a personnel connection to someone working for the organization. 26 Figure 2: Primary Channels by which Donors and Volunteers Learned about FCF Cambodia Organizations Respondents who donated to and/or volunteered with CCT, FI, and RO also had a tendency to volunteer domestically as well—63 percent of respondents (54 out of 86) reported having volunteered for a domestic organization in the country where they live. As shown in Figure 3, of those, 67 percent (36 of 54) reported that they found out about the organization they volunteered with through word of mouth, and 17 percent (9 of 54) reported they found out about the opportunity through their job. Several less common channels are also shown in Figure 3. 38% 28% 7% 6% 4% 4% 4% 3% 2% 2% 1% 1% Television A Friend Personal connection with someone from the organization Online search for organization to donate to An Event Restaurant Poster Gift shop Online advertisement Online research an on issue Social Media Newspaper 27 Figure 3: Channels by which Respondents Learned about Domestic Organizations with Which to Volunteer Twenty percent of respondents (17 out of 86) also reported that they had volunteered with other international organizations (outside of an FCF Cambodia organization for those who had volunteered with CCT, FI, or RO), and of those 47 percent (8 out of 17) indicated they had volunteered with an RCI or orphanage. The ways they found out about the international organizations were nearly identical to those reported as channels for domestic volunteering, with word of mouth and their job listed as the most popular and second most popular channels. Positive deviant interview respondents reported learning about the organization they initially donated to and/or volunteered with through similar channels. Four out of ten respondents learned about the organizations through word of mouth and/or personnel connections, while three out of ten respondents learned of them through news media such as television. One respondent learned of an organization through radio, and another learned of an organization through a travel/volunteer placement agency. Positive deviant interview respondents also reported learning more about these organizations through the internet after having been introduced through other means. Two out of ten positive deviant respondents worked for volunteer organizations or Cambodian nonprofits providing non-residential care, and weren’t donors themselves. Evidence from the literature differs somewhat from these survey and interview results, finding that the predominant channels for recruiting volunteers include: online search optimization (where links to RCIs or RCI-placement hubs appear in response to a general search for volunteering opportunities); promotion by volun-tourism agencies, RCIs, and peers; and direct outreach by local agents in tourist venues—hotels, travel agencies, tourism destinations, tuk tuks (BCN 2014). A recent study commissioned by UNICEF interviewed representatives from 18 RCIs in Cambodia confirms some of these findings. The study found that most of the donors to those RCIs had been “recruited” through online campaigns, networks created by former volunteers, and local and international travel agencies (EMC 2016). The new research presented here likely does not capture this last group because it is based on online outreach—likely capturing respondents who also spend time online rather than as tourists at hotels, hostels, and tourist agencies in Cambodia. A study carried out by UNICEF (2011) 55% 14% 8% 6% 5% 3% 3% 3% 1% 1% 1% Word of mouth Job Telephone Email Internet Search Television Ad Newspaper Ad Volunteer Bank Online Ad School Church 28 found that about 80 percent of tourists in Siem Reap who had considered volunteering for RCIs had not made any concrete plans to do so prior to arriving in Cambodia. Recently, Better Care Network14 published a study that sought to identify key words volunteers use in searching online for organizations with which to volunteer. They found that for every search for “volunteering in orphanages,” there were about 400 searches for “volunteering abroad,” suggesting that most searches conducted by potential volunteers are for general opportunities, not for opportunities tied to orphanages in particular. And, yet, 20 out of the first 23 organizations listed in Google results from searching “volunteering abroad” featured RCIs or links to RCI placement webpages (BCN 2014). Channels to Discourage Support for RCIs As shown in Figure 4, when asked to rank the most effective ways to disseminate messages discouraging support for RCIs, survey respondents reported the top five channels would likely be through (1) social media, (2) television, (3) online advertisements,15 (4) email, and (5) radio. Social media ranked highest on this list of potential communication channels, with 45 percent of respondents ranking it as the best way to get information out. Television advertisements ranked second, with 12 percent of respondents ranking it as the best way to share information with potential donors or volunteers. Online advertisements ranked third, with 14 percent of respondents ranking it as the best way to get information out. Email ranked fourth, with 6 percent of respondents ranking it highest. Radio ranked fifth, with 1 percent of respondents ranking it as the best way to share this type of information. Figure 4: Recommended Channels for Distributing Information about RCIs 14 An international network of organizations committed to supporting children without adequate family care around the world. 15 Since the survey was disseminated online, it undoubtedly was responded to by people unusually favorable to online communication channels. 0 5 10 15 20 25 30 35 40 Social Media Television Ad Online ad Radio Ad Email Phone Campaign Other Best Mode Second Best Mode Third Best Mode 29 In trying to target potential SBCC campaigns to the right audiences, it is important to identify differences in potential audiences. We found that certain demographic characteristics correlated with respondents’ preferences among these channels. People with advanced degrees tended to prefer email more than people with primary degrees. People with no children or no religious affiliation were more likely to prefer television ads than people with children or religious affiliations, respectively. Those with a religious affiliation tended to prefer phone campaigns more frequently than did those without a religious affiliation. Positive deviant interviewees most commonly suggested that social media (especially for younger generations—four out of ten respondents), mass media (including television shows, advertisements, or the news—four out of ten respondents), or online ads (two out of ten respondents) would be the best way to reach potential donors and volunteers with messages. Two respondents also suggested specifically targeting tourism centers, and another suggested targeting faith-based groups. Male respondents (all three out of ten total respondents) were more likely to mention personal connections as an important factor in messaging. Despite the fact that both survey and positive deviant respondents suggested that social media would be the best way to target potential donors or volunteers with key messages about the harms of RCIs, respondents typically did not talk about social media as a source of their own information about potential donation and volunteer opportunities. One respondent specified that while she does not use social media to learn about the details of organizations, social media might initially call her attention to an organization that she then researches more online. Survey respondents and positive deviants cited some communication channels as particularly effective in convincing them not to donate to RCIs, including CCT’s articles and TED talks, talks given by organization representatives, personal communication with representatives from FCF Cambodia TSG organizations, and presentations given by FI. Targeting Potential Donors and Volunteers Based on the literature, the volunteers working in RCIs in Cambodia come mainly from Australia, New Zealand, Great Britain, England, Germany, and France. The same is true of donors. Volunteers are also likely to come from middle- to upper-class families, are more likely to be women than men, and likely include many students and Christians (Brown 2005; McGehee 2012; Mostafanezhad 2013; Guiney 2014; Reas 2013; van Doore 2016; BCN 2014). Though it could just be a matter of bias in who chose to respond to our survey, the survey results corroborated some of these hypotheses. For instance, we found that 73 percent of respondents were women, and 79 percent were from Australia. However, we also found that respondents were more likely to be old than young and were more likely not to be religiously affiliated than to be religiously affiliated. 30 CONCLUSIONS COMMUNITY MESSAGES QUESTION 1—WHAT TYPES OF SBCC CAMPAIGN MESSAGES ARE LIKELY TO RESONATE WITH CAREGIVERS TO CHANGE THEIR ATTITUDES AND PERCEPTIONS ABOUT RCIS? Findings overwhelmingly pointed to two challenges that communities face, primarily with ensuring adequate education for their children and secondarily with issues of poverty. While caregivers want to keep their children at home, their aspirations for their children often supersede those desires, as they believe children receive a better education in RCIs. And, caregivers recognize that education is essential for children to find non-agricultural jobs and be successful. However, some feel they do not have adequate access to educational facilities, like secondary schools and they can’t afford school fees. They also feel incapable of supporting their children’s studies, believing children will study harder in RCIs than at home. Caregivers feel pressure to support their families’ basic needs and often cannot make enough money farming to do so. Given that farming is the predominant livelihood activity in much of Cambodia, caregivers often must migrate to earn a sustainable income. This means children are left with other caregivers if families can’t afford to take them with them, or if families do not have those backup supports, the children are placed in an RCI. While many study participants suggested that messages focused on keeping families together would likely resonate with community members, many respondents made it clear that they understood the importance of keeping their families together but could not do so because of other pressures. On the other hand, the vast majority of community members are unfamiliar with the harms of many of the RCIs and overwhelmingly believe RCIs to be positive additions to their communities that serve an important need. Based on this information, the following messages might resonate with community members:  Messages that provide information about educational and financial support available in communities through the HFP model or otherwise  Messages that show that children who are placed in RCIs are less likely to find educational, career, and financial success than their peers  Messages that explain the harms of RCIs with regards to abuse, developmental issues, etc. Regardless of the message, any SBCC campaign implemented should also ensure inclusion of information on other ways for families to access services or get the childcare they need to ensure any SBCC campaign does not results in negative effects on families. COMMUNITY COMMUNICATION CHANNELS QUESTION 2—WHAT COMMUNICATION CHANNELS ARE LIKELY TO MOST EFFECTIVELY REACH THE GREATEST NUMBER OF CAREGIVERS? Findings demonstrate that village chiefs are clearly important information gatekeepers and decision makers in communities. By and large, community members prefer personal communication channels, and they trust village chiefs. Village chiefs typically distribute information via community meetings, which is the channel of communication most FGD participants said they would prefer. However, the fact that 31 they have sometimes been involved with recruiting children to RCIs means that if CCT decides to target village chiefs to disseminate information about RCIs, they may face a steep curve in changing perceptions and incentive structures. Next, community members reported having access to mobile phones and television, and literature corroborated this information. Community members also suggested that they often look to their cell phones and television, especially the news, for information. Television, however, can be expensive, and it can be difficult to target messages to specific audiences. On the other hand, literature showed that while mobile phone-based SBCC interventions have had mixed success, initial evidence suggests that voice￾message interventions have worked in Cambodia. As such, the following channels are likely to effectively reach the most caregiver (in priority order):  Community meetings held by village chiefs  Interactive voice messaging via cell phones  Television ads, likely played during the news DONOR AND VOLUNTEER MESSAGES QUESTION 1—WHAT TYPES OF SBCC CAMPAIGN MESSAGES ARE LIKELY TO RESONATE WITH POTENTIAL RCI DONORS AND VOLUNTEERS TO CHANGE THEIR ATTITUDES AND PERCEPTIONS ABOUT RCIS? Evidence from literature and survey and KII respondents suggests that while most donors and volunteers are well-meaning, they do not realize that many of the children who are placed in RCIs are not orphans. Respondents said that recognizing that they themselves did not realize this prior to becoming involved with FCF Cambodia member organizations. They also provided lots of good fodder for possible messages, including “Orphanages create orphans.” Further, one respondent suggested sharing the percentage of children who are not orphans in Cambodian RCIs as a telling data point. Evidence suggests that messages focused on this theme or orphanages creating more orphans is likely to resonate with the most potential donors and volunteers. The second theme most likely to resonate with potential donors and volunteers is one that describes the harms of RCIs. Donors and volunteers mentioned the following more specific themes within this general topic area: exploitation, revenue-generation, and developmental issues for children. Literature also showed that half of Cambodian RCI volunteers are not aware that RCIs can be seriously detrimental to the long-term development of a child, or that they are run as for-profit businesses that must rely on the constant flow of children for their own survival. The third most likely message to resonate with potential donors and volunteers is a message focused on the importance of keeping families together. Donors and volunteers suggested that messages about families are the reason why they chose not to donate to or volunteer with an RCI. One survey respondent said, “I believe that every child has a right to family.” Donors and volunteers also suggested that no matter what the message, it must include information about how donors and volunteers can help rather than just telling them what not to do (re: not donating to or volunteering with RCIs). 32 DONOR AND VOLUNTEER CHANNELS QUESTION 2—WHAT COMMUNICATION CHANNELS ARE LIKELY TO MOST EFFECTIVELY REACH THE GREATEST NUMBER OF POTENTIAL DONORS AND VOLUNTEERS TO RCIS? While respondents overwhelming recommended social media as the best channel by which to reach potential donors and volunteers with a message about the harms of RCIs, very few of them said they found out about the organizations they donate to or volunteered with through social media. On the contrary, they said they learned about the organizations through personal connections and word-of￾mouth. However, it would be difficult for FI to disseminate information through word-of-mouth. The organization has much more power to affect social media, and social media could be considered one of the most common ways we make personal connections and spread information through word-of-mouth today. Next, many survey respondents learned about the domestic and international organizations they donated to or volunteered with through television, and many also recommended television as the best channel for disseminating information about RCIs to potential donors/volunteers. Online searches came up as the channel by which about five to six percent of survey respondents learned about organizations they donated to or volunteered with. And, literature suggested that this was an even more common method and one that was highly recommended by survey respondents to get information out to potential donors and volunteers. Finally, while they were not mentioned very frequently among survey or KII respondents, hotels, hostels, and travel agencies were shown to be a prominent channel for getting information out in literature. And, it is likely that selection bias issues with our sampling and data collection methods caused this channel not to come up as often in our research. 33 Recommendations COMMUNITIES While there is variation in both potential messages and potential channels by which to reach community members, conducting an experiment that varies both will require a very large sample size and will be extremely resource intensive. As such, we recommend an experiment that either varies the message or the channel but not both. While we indicate three potential channels in our conclusions, village chiefs clearly came out as the likely best channel of communication and are also likely the least resource-intensive channel. Further, it is clear that caregivers by-and-large understand that children are better off at home, but familial challenges and corresponding incentive structures make it difficult for caregivers to always keep children at home. So, not all messages are likely to have the same effect on caregivers. It will be important to identify which ones flip the incentive structure. As such, we recommend a community SBCC campaign in which CCT varies three messages focused on:  Educational and financial support available in communities  The relative less success that children who are placed in RCIs are likely to find in terms of their education, career, and financial sustainability than their peers  The harms of RCIs with regards to abuse, developmental issues, etc. These should be distributed by village chiefs via community meetings. All messages should also include information on ways for families to access services or get the childcare they need to ensure there are no unintended negative externalities of the campaigns. DONORS AND VOLUNTEERS While there is also variation in both potential messages and potential channels by which to reach potential RCI donors and volunteers, conducting an experiment that varies both will require a very large sample size and will be extremely resource intensive. As such, we recommend an experiment that either varies the message or the channel but not both. Further, while FI does not have a clear message theme that they want use to target potential donors or volunteers, the content of the message does not seem likely to matter as much as the channel by which the information in disseminated. The reason for this is that donors and volunteers want to do the right thing, and any indication that they are not doing the right thing will make them likely to want to change their behavior. However, figuring out how to target people based on their potential future actions (e.g., donating to or volunteering with an RCI) can be difficult, as the future is unknown. Thus, we recommend a campaign that varies the channel of disseminating a message—likely one about orphanages creating orphans—using social media; online advertisements based on key-word searches; and/or flyers and posters placed in hotels, hostels, and travel agencies in Cambodia. These are preliminary recommendations for Feedback Experiments. The RF MERL Consortium will follow this report with a formal Feedback Experiment Options Memo, which we will discuss with CCT and FI before finalizing the Feedback Experiments. 34 ANNEXES ANNEX I: DATA COLLECTION INSTRUMENTS KEY INFORMANT INTERVIEW: VILLAGE LEADERS GUIDE Introduction: 1. Please introduce yourself by name, age, village leadership position, and the roles and responsibilities of your position. 2. How many families live in this village? 3. Please tell me about your village and the families that live here. Respondent Role in the Community 4. What type of interaction do you have with families in your village? 5. What type of influence do you have on families in your village? 6. Who do you go to for support when making decision about the village and families here? Family Structure: 7. In your village, which family members normally live together in a household? 8. What are other examples of how families live together? Hopes for Children: 9. What hopes do you have for children in your community? Potential Probes: What do you hope their lives are like in the future? What do you hope they will achieve in life? 10. What hopes do you think their families have for them? 11. How likely is it that children will reach/achieve those hopes? Why? 12. What, if anything, can be done to ensure children are able to reach/achieve those hopes? Caring for Children: 13. Who normally takes care of children in your community? 14. What happens when the normal caregivers are not available? 15. What are the other ways your community takes care of children? 16. How do families make decisions about caring for children in your community? Family Challenges: 17. Now, I would like to talk about challenges, not just those related to caring for children but all community challenges. What are the greatest challenges families in your community face? 35 Potential Probes: If they can’t come up with any, you could say, what about education, healthcare, nutrition, job training, safety, domestic violence, drug abuse, death? Why do these challenge exist? Can you expand on that? If they say education, health, nutrition, job training, or safety, ask: If _______ is a challenge, what are the specific challenges with ________ offerings in your community? (fill in blanks with problems identified by the respondent – i.e. – education, health, nutrition, job training, safety, etc.) If they say domestic violence or drug abuse or death of caregivers, ask: Why is that a major issue in your community? How might these problems be avoided or better addressed in your community? 18. What other challenges do they face, if any? 19. How do these challenges affect children in the household? Community Support and Services: 20. How do community members support each other when they face these challenges? 21. What types of services and resources are available in your community to help families face these challenges? 22. To what extent do families know about these services and resources? a. If they don’t know much about them, why not? 23. How do families typically learn about resources in the community? 24. What recommendations do you have for getting the word out about services in your community? 25. What, if any, additional resources and services do you think are needed in your community? Why? Modes of Communication 26. About what percentage of families in your community have access to: a. Television b. Radio c. A mobile phone d. A newspaper e. The internet f. Facebook g. Other social media 27. Approximately how many hours per week do members of the community consume each of these types of media? a. Television b. Radio c. A mobile phone d. A newspaper 36 e. The internet f. Facebook g. Other social media 28. How do people in your community typically receive information? 29. How do they get information about services to take care of children or support their children? 30. What sources do they trust for information about their children? 31. What sources do they not trust? 32. What do you think is the best way to get an important message out to the families in your community? 33. Will all families receive the message if it is delivered in that way? If not, how could we ensure we reach all families? RCIs 34. Are there RCIs in your community (If they have already mentioned that there are, you do not need to ask this question – but be sure you know if the RCIs are actually in their community) If no, skip to 37 35. What types of RCIs are there? 36. What role do they play in your community? Skip to 38 37. Do RCIs in other communities have any effect on families in your community? 38. What is your impression of the RCIs? 39. How frequently do children from your community get placed in an RCI? Why? Potential Probes: What are families’ reasons for placing children in RCIs? What types of problems lead them to place children in RCIs? Are there particular events or severe problems that precede placing a child in RCIs? Do families who place children in RCI differ from typical families in your community? 40. What are the characteristics and conditions of children placed in an RCI? How do these characteristics differ from those of other children in your community? 41. Can you think of a child from your community that has been placed in an RCI? If so, can you describe his/her family situation, why he/she was placed there, and who made the decision to place him/her? 42. How common is it that more than one child from the same household is placed in an RCI? Why? 37 43. Who is involved in the decision about whether a child gets placed in an RCI? Potential Probes: Do you have any control over the decision of whether a child is placed in an RCI? Does anyone else in your community take part in decision of whether to place a child in an RCI? 44. What are the benefits of RCIs, if any? 45. What are the harms of RCIs, if any? 46. What would you change about RCIs, if anything? 47. Have you be in contact with or interacted with RCIs in your community or elsewhere? What kinds of interactions have you had with them? 48. Is there anything else you think I should know about RCIs? Framing Behavior Change Messages 49. A local nonprofit organization is planning to create some messages with the goal of helping to keep families together. In thinking about that goal, what messages do you think would most resonate with families in your community? 50. Do you think any of the following would resonate? a. Messages focused on the importance of families staying together and the benefits children who grow up in a family receive? b. Messages focused on the negatives of children not living with their family? c. Messages focused on ways of addressing some of the challenges you identified earlier? d. Messages focused on other available services and resources in your community? If he/she selects this one, ask: Which resources and why? e. Messages focused on the negatives of RCI care? 51. How should these messages be delivered to families to have the most impact? 52. Only ask if the interviewee has mentioned some families sending children to RCIs - To what extent do you think such messages will work to stop families from sending children to RCIs? Final Thoughts 53. Is there anything we haven’t covered that you think I should know? 54. Do you have any questions for me? 55. Is there anyone else you think I should speak with about these topics? FOCUS GROUP GUIDE: CAREGIVERS Icebreaker Activity – If you are working with a group who do not know each other well, have everyone work with a partner and get to know the other person. Then the partner groups will introduce each other saying their partner’s name/pseudonym, occupation, age, who they live with, and number and age of children in their household (including how many are 38 still at home). To make it more fun, you can also ask them to come up with one unique thing they have in common. If they do know each other well, just ask them to introduce themselves, their occupation, age, who they live with, the number and age of children in their household (including how many are still at home) [10 minutes] Topic – Family Structure (personal and community) In your village, which family members normally live together in a household? (Use information from introduction to prompt the group, if necessary. For example, are other families in your community similar to what we’ve heard described? Probe for discussion among members.) Follow up on stories or opinions expressed previously as necessary or appropriate. topics are optional; you do not need to ask each question listed here. Potential probes: ● Who usually makes important decisions in the household? ● What are other examples of how families live together in your community? ● Are there extended family groups who live in the village? How are they related? ● Tell us more about your family. Are there roles that certain members of a family play? ● Tell us about how your family cares for children. Topic – Caring for Children Who generally cares for children in your household? Potential probes: ● How does the childcare situation in your household differ from that of other households in your community? ● What responsibilities do different caregivers have in your household? ● Do responsibilities of who cares for children change as children age? ● Who cares for the children when the typical caregiver needs to be away? ● What are other ways your community takes care of children? ● How does your community choose which ways to care for children? Hopes for your Children What hopes and dreams do you have for your children? Potential probes: ● What do you hope they will be like when they grow up? Are your hopes different for boys or girls? o What kind of job do you hope they will have? o What things do you do to try to help them reach those goals? ● To what extent do you feel happy with the way you are raising them now? Do you feel the same about raising both boys and girls in your family? o What things, if any, do you wish you could do differently in raising them? Topic - Challenges and Obstacles Now, I’d like to move to talking about challenges your family faces in general, not just related to child care. What are some of the biggest obstacles/challenges that families face here? Potential probes: ● Can you expand on how this obstacle/challenge affects children, if it does? ● Tell me more about why this is challenge in your community. 39 ● Do children have access to education in your community? Please explain. Is this ever a challenge? ● Do children have access to adequate health services in your community? Is this ever a challenge? ● Do households have access to enough nutritious foods? What other kinds of problems in households or communities negatively affect children (for example, are domestic violence, drug abuse, or other family issues a problem)? Please explain. Topic – Service and Support Availability What types of support do community members provide to help families facing these challenges? What services are available in your village to help families facing these challenges? Potential probes: ● Do you know of families who use these resources? o If so, why those resources? Why do families choose those options over other options? o Where do families learn about these resources? Topic – Communication Channels What types of media do you use on a regular basis (please include information about watching TV, listening to the radio, reading newspapers, accessing information on mobile phones, etc.)? Potential probes:  If they mention TV, ask: What TV channels do you watch most often?  If they mention radio, ask: What radio stations do you listen to most often?  If they mention mobile phones, ask: What mobile phone applications do you use most often? Do most people in your community have access to TVs, radios, newspapers, and mobile phones? Please tell me about each one. Where do you find out about important information in your community? Potential probes: ● Is this the same source you would use for information on community services to support your children? If not, what source would you use for that information? ● Is that how you prefer to get information? o If not, how would you prefer to get information about services available in your community? o If so, why? ● What sources of information do you most trust and why? ● What sources of information do you least trust and why? Topic – RCIs’ role in the Community Transition if RCIs have been mentioned: Can you tell me how you define an RCI? I would like to hear more about the role of RCIs and how they’re viewed in your community. Transition if RCIs have NOT been mentioned: When thinking about the different services that exist in your community (recap what has been listed or discussed), how do RCIs fit in? How do you define RCIs? Potential probes:  Please name a few of the RCIs in your community or a neighboring community. 40  What do you know? ● How did you learn about this? ● How are RCIs perceived by you/people like you? ● Have you (or people you know) interacted with an RCI? o Under what circumstances? ● What are the benefits or harms of RCIs? Topic – Why Caregivers Place Children in RCIs Transition if families placing children in RCIs have been mentioned: Why do families choose to place their children in RCIs? Transition if families placing children in RCIs have NOT been mentioned: I have heard from others in similar communities that families sometimes place their children in RCIs. Why do you think they do this? Potential probes:  Do you think this is the right choice for families? o Why or why not? o When or when not? ● Under what circumstances do families place children in RCIs? Why do they make that choice rather than accessing other community resources or services? o Does access to a safe living environment or education, health, nutrition, or job training services have a role? Please explain. o How, if at all, are these circumstances different from the challenges families typically face? ● What are the characteristics of children placed in RCIs? Why do these characteristics lend themselves to placement in an RCI? o How, if at all, do they differ from other children in your community? o If a family chooses to place a child in an RCI, do they place all of their children, or just one or two?  How do they make the decision of which child/children to place? Topic – Possible RCI Messaging What types of messages do you think would most resonate with families in your community in an effort to keep children living with their families? Potential probes (types of messages in this case): ● Messages about the importance of keeping children living with their family and the benefits of families staying together ● Messages about the negatives that occur when families are split up ● Messages focused on other education opportunities in the community ● Messages focused on the negatives of RCIs ● Others? KEY INFORMANT INTERVIEW: CCT DONORS AND VOLUNTEERS GUIDE Demographic Information: 1. How old are you? 2. Where were you born? 41 3. Where do you currently live (city and state in the US or city and country outside the US)? 4. What is the highest level of education you have completed? 5. What is your employment status (e.g., are you full time, part time, unemployed, retired, etc.)? 6. What is your approximate annual household income? I am happy to provide ranges for you. If he/she says yes, ask: Can you estimate the range it falls into: i. $0 ii. $1 – $10,000 iii. $10,001 - $30,000 iv. $30,001 - $50,000 v. $50,001 - $75,000 vi. $75,001 - $100,000 vii. $100,001 - $150,000 viii. $150,001 - $250,000 ix. $250,001+ x. Don’t know xi. Choose not to respond Also, be sure to clarify the currency they are referring to. 7. What is your occupation? 8. What is your religious affiliation? 9. Do you have children? If so, how many? 10. Have you visited Cambodia before? Donation or Volunteer Information: 11. What is your impression of Cambodian orphanages and residential care institutions? 12. How did you become involved with [CCT/organization name]? 13. What made you decide to donate to/volunteer with [CCT/organization name]when it was an orphanage/ residential care institution (Residential care refers to group living arrangements for children without primary caregivers or whose biological parents are unable to care for them. It is meant to provide 24-hour care by paid staff, meeting children’s basic needs of shelter, food, clothing and education. These can be places of safety for emergency care, transit centres in emergency situations, and all other short and long-term residential care facilities including group homes. Other terms used for this form of care are institutional care and orphanages)? 14. What was your impression of [CCT/organization name] when it was an orphanage/residential care institution? 15. To what extent do you think [CCT/organization name] was like other orphanages/residential care institutions in Cambodia? 42 16. Do you still donate to/volunteer with [CCT/organization name] today, now that it has changed its care model? a. If yes, why did you decide to continue donating to/volunteering with them even though they changed their model? i. Was there a specific instance, interaction, piece of information, or marketing campaign that made you decide to continue donating to/volunteering with [CCT/organization name]? If so, please describe it and why it was so effective. b. If no, why did you decide to stop donating to/volunteering with them? i. Was there a specific instance, interaction, piece of information, or marketing campaign that made you decide to stop donating to/volunteering with [CCT/organization name]? If so, please describe it and why it was so effective. 17. Would you recommend that your friends consider donating to or volunteering with [CCT/organization name]? Why or why not? Other Donation and Volunteer Information: 18. Do you donate to/volunteer with any Cambodian orphanages or residential care institutions now? Why or why not? 19. Would you recommend that your friends consider donating to or volunteering with another Cambodian orphanage or residential care institution? Why or why not? Modes of Communication 20. What do you think would be the best way to get a message out to potential donors/volunteers about the harms of orphanages or residential care institutions? 21. In thinking about a campaign to discourage donations to or volunteering with Cambodian residential care institutions, what type of message do you think would resonate with potential donors/volunteers? 22. Where do you typically get information on organizations you might want to donate to or volunteer with? 23. How often do you get emails from [CCT/organization name]? 24. How likely are you to read those emails? 25. What types of subject lines, email layouts, or other key features make you more likely to read one email from [CCT/organization name] than another? Final Thoughts 26. Is there anything we haven’t covered that you think I should know? 27. Do you have any questions for me? 28. Is there anyone else you think I should speak with about these topics? 43 KEY INFORMANT INTERVIEW: OTHER ORGANIZATION DONORS AND VOLUNTEERS GUIDE Demographic Information: 1. How old are you? 2. Where were you born? 3. Where do you currently live (city and state in the US or city and country outside the US)? 4. What is the highest level of education you have completed? 5. What is your employment status (e.g., are you full time, part time, unemployed, retired, etc.)? 6. What is your approximate annual household income? I am happy to provide ranges for you. If he/she says yes, ask: Can you estimate the range it falls into: i. $0 ii. $1 – $10,000 iii. $10,001 - $30,000 iv. $30,001 - $50,000 v. $50,001 - $75,000 vi. $75,001 - $100,000 vii. $100,001 - $150,000 viii. $150,001 - $250,000 ix. $250,001+ x. Don’t know xi. Choose not to respond Also, be sure to clarify the currency they are referring to. 7. What is your occupation? 8. What is your religious affiliation? 9. Do you have children? If so, how many? 10. Have you visited Cambodia before? Donation or Volunteer Information: 11. What is your impression of Cambodian residential care institutions or orphanages? 12. How did you become involved with [FI/organization name]? 13. What made you decide to donate to/volunteer with [FI/organization name]? Have you ever donated to a residential care institution or orphanage? Why or why not? If not, skip to 16 44 14. Have you ever donated to a Cambodian residential care institution or orphanage? Why or why not? 15. Do you currently donate to/volunteer with any Cambodian orphanages or residential care institutions? Why or why not? 16. Only ask if relevant – Why did you stop donating to/volunteering with the Cambodian orphanage/residential care institution? a. Was there a specific instance, interaction, piece of information, or marketing campaign that made you decide to stop donating to/volunteering with the Cambodian orphanage/residential care institution? If so, please describe it and why it was so effective. 17. Would you recommend that your friends consider donating to or volunteering with a Cambodian orphanage or residential care institution? Why or why not? Modes of Communication 18. What do you think would be the best way to get a message out to potential donors/volunteers about the harms of orphanages or residential care institutions? 19. In thinking about a campaign to discourage donations to or volunteering with Cambodian residential care institutions, what type of message do you think would resonate with potential donors/volunteers? 20. Where do you typically get information on organizations you might want to donate to or volunteer with? 21. How often do you get emails from [FI/organization name]? 22. How likely are you to read those emails? 23. What types of subject lines, email layouts, or other key features make you more likely to read one email from [FI/organization name] than another? Final Thoughts 24. Is there anything we haven’t covered that you think I should know? 25. Do you have any questions for me? 26. Is there anyone else you think I should speak with about these topics? DONOR AND VOLUNTEER SURVEY [Note below does not show skip patterns] Question 1 This survey is being sent to you by [organization name] on behalf of Results for Development (R4D), Abt Associates, and Mathematica, three US-based organizations that have formed a consortium to study residential care institutions (RCIs) in Cambodia. The study is funded by the United States Agency for International Development (USAID), and the results of the study will be used to determine how best to reach potential volunteers and donors to provide them with more information about the Cambodian organizations they are donating to or volunteering with. We would greatly appreciate your participation in the following survey, which includes questions about your history of charitable giving and volunteering. Your participation is entirely optional. 45 The survey will take about 10 minutes to complete. You are free to skip any questions you do not want to answer. If you choose to skip a question, just mark the “choose not to respond” option. Your participation in this study is entirely confidential. While we ask you to provide your name, doing so is entirely optional. If you do provide it, we wi ll only use the information to connect your survey back to other demographic information about you from your profile with [organization name], Whether you provide your name or not, your identity will be kept confidential among the research team. Any information shared from this discussion with USAID or others will be reported in aggregate. When a quote is used, it will not include your personal identifying information. Per USAID requirements, the data we collect here today will be uploaded to USAID’s data warehouse for use by authorized researchers; however, all identifying information will be removed. I understand, certify that I am over the age of 18, and agree to participate I do not agree to participate, or am not 18 years old Question 2 What is your name? Question 3 What is your sex? Male Female Other Choose not to respond Question 4 What is your age? Question 5 Where were you born? --- Question 6 Where do you currently live? --- Question 7 What is the highest level of education you have completed? Less than primary/elementary school Primary/elementary school Some secondary school/high school, no diploma Secondary school/high school graduate, diploma or the equivalent (i.e. GED) Some college credit, no degree Trade/technical/vocational training Associate’s degree or other 2-year technical degree Bachelor’s degree Master’s degree 46 Professional degree Doctorate degree Choose not to respond Question 8 What is your marital status? Single, Never Married Married Living with partner Separated Divorced Widowed Choose not to respond Question 9 Do you have any children? Yes No Choose not to respond Question 10 What is your employment status? Employed Full-Time Employed Part-Time Self-employed Not employed, but looking for work Not employed and not looking for work Homemaker Retired Student Other, please describe Choose not to respond Question 11 What is your annual household income (in USD)? Under $20,000 $20,000 - $30,000 $30,000 - $40,000 $40,000 - $50,000 47 $50,000 - $75,000 $75,000 - $100,000 $100,000 - $150,000 $150,001 - $250,000 More than $250,000 Don't know Choose not to respond Question 12 What is your religious affiliation? Mainline Protestant Evangelical Catholic Mormon Jehovah’s Witness Other Christian Jewish Buddhist Hindu Muslim Atheist Agnostic Other, please describe Don’t know Choose not to respond Question 13 Have you visited Cambodia before? Yes No, but I am planning to soon No, and I have no plans to visit Choose not to respond Question 14 How did you become involved with the [organization name]? I met someone from the organization at an event A friend told me about the organization I saw an advertisement for the organization online and clicked on it 48 I heard about the organization on social media I was researching an issue and found it I was searching for an organization to donate to and found it I was searching for an organization to volunteer with and found it Other, please describe Don’t know Choose not to respond Question 15 Have you ever donated money to [organization name]? Yes No Choose not to respond Question 16 What made you decide to donate to [organization name]? Question 17 How did you find out you could donate to [organization name]? Through my job Through word of mouth (e.g., a friend, relative, etc.) The organization called asking for a donation The organization emailed asking for a donation I saw a television ad asking for donations I saw an online ad asking for donations I found the organization while browsing the web and decided to donate I heard a radio ad asking for donations Other, please describe Don’t know Choose not to respond Question 18 Have you ever donated to any other organization(s)? Yes No Choose not to respond Question 19 What type(s) of domestic organization(s) have you donated to, if any? Select all that apply. 49 Animal charities Arts and culture charities Disaster relief and humanitarian NGOs Environmental charity (including parks, nature conservatories, and forests) Health charity (e.g., organizations that support and treat the sick and disabled, work on cures for diseases, and promote public awareness of specific health risks; hospitals; etc.) Homeless shelter, soup kitchen, or other service to help the homeless Local service organization (e.g., the fire department, police department, etc.) Peace and human rights organizations (e.g., ACLU) Religious organizations and/or churches University or alumni group Women’s rights organization (e.g., Planned Parenthood, domestic abuse organizations or shelters, etc.) Youth organization (e.g., big brothers/sisters, boys and girls club, etc.) Other, please describe None, I have only donated to international organizations Don’t know Choose not to respond Question 20 What other type(s) of international organization(s) have you donated to, if any? Select all that apply Animal charity Arts and culture charities Child sponsor organizations Disaster relief and humanitarian NGOs (e.g., Red Cross, Save the Children) Environmental charity (including parks, nature conservatories, and forests) Health charity (e.g., organizations that support and treat the sick and disabled, work on cures for diseases, and promote public awareness of specific health risks; hospitals; etc.) Peace and human rights organizations (e.g., Amnesty International, Human Rights Watch, etc.) Orphanages or residential care facilities Religious organizations and/or churches University or alumni group Youth organization (e.g., big brothers/sisters, boys and girls club) 50 Other, please describe None, I have only donated to domestic organizations and [organization name] Don’t know Choose not to respond Question 21 Would you ever consider donating to an orphanage or residential care institution? (Residential care refers to group living arrangements for children without primary caregivers or whose biologica l parents are unable to care for them. It is meant to provide 24-hour care by paid staff, meeting children’s basic needs of shelter, food, clothing and education. These can be places of safety for emergency care, transit centres in emergency situations, an d all other short and long-term residential care facilities including group homes. Other terms used for this form of care are institutional care and orphanages.) Yes No Choose not to respond Question 22 Would you ever consider donating to an orphanage or residential care institution in Cambodia? Yes No Choose not to respond Question 23 Why or why not? Question 24 Have you ever volunteered with a domestic nonprofit or charitable organization? Yes No Choose not to respond Question 25 What type(s) of domestic organization(s) did you volunteer with? Select all that apply Animal charities Arts and culture charities Disaster relief and humanitarian NGOs Environmental charity (including parks, nature conservatories, and forests) Health charity (e.g., organizations that support and treat the sick and disabled, work on cures for diseases, and promote public awareness of specific health risks; hospitals; etc.) Homeless shelter, soup kitchen, or other service to help the homeless Local service organization (e.g., the fire department, police department, etc.) Peace and human rights organizations (e.g., ACLU) Religious organizations and/or churches 51 University or alumni group Women’s rights organization (e.g., Planned Parenthood, domestic abus e organizations or shelters, etc.) Youth organization (e.g., big brothers/sisters, boys and girls club, etc.) Other, please describe None, I have only donated to international organizations Don’t know Choose not to respond Question 26 What made you decide to volunteer with the organization(s)? Question 27 How did you find out you could volunteer with the organization(s)? Select all that apply Through my job Through word of mouth (e.g., a friend, relative, etc.) The organization called asking for volunteers The organization emailed asking for volunteers I saw a television ad asking for volunteers I saw an online ad asking for volunteers I found the organization while browsing the web and decided to volunteer I heard a radio ad asking for volunteers Other, please describe Don’t know Choose not to respond Question 28 Have you ever volunteered with an international nonprofit or charitable organization? Yes No Choose not to respond Question 29 What type of international organization(s) did you volunteer with? Select all that apply Animal charity Arts and culture charities Child sponsor organizations Disaster relief and humanitarian NGOs (e.g., Red Cross, Save the Children) 52 Environmental charity (including parks, nature conservatories, and forests) Health charity (e.g., organizations that support and treat the sick and disabled, work on cures for diseases, and promote public awareness of specific health risks; hospitals; etc.) Peace and human rights organizations (e.g., Amnesty International, Human R ights Watch, etc.) Orphanages or residential care facilities Religious organizations and/or churches University or alumni group Youth organization (e.g., big brothers/sisters, boys and girls club) Other, please describe None, I have only donated to domestic organizations and [organization name] Don’t know Choose not to respond Question 30 What made you decide to volunteer with the organization(s)? Question 31 How did you find out you could volunteer with the organization(s)? Select all that apply Through my job Through word of mouth (e.g., a friend, relative, etc.) The organization called asking for volunteers The organization emailed asking for volunteers I saw a television ad asking for volunteers I saw an online ad asking for volunteers I found the organization while browsing the web and decided to volunteer I heard a radio ad asking for volunteers Other, please describe Don’t know Choose not to respond Question 32 Would you ever consider volunteering with an orphanage or residential care institution? Yes No Choose not to respond Question 33 53 Why or why not? Question 34 What is your impression of Cambodian orphanages or residential care institutions? I do not know enough about them to have an impression I think they provide a good service I think they are not an ideal solution, but it is better than other alternatives I think they are good for some children but harmful for other children I think they are harmful /disruptive to families Other, please explain Don’t know Choose not to respond Question 35 Would you recommend that your friends or family consider donating to or volunteering with a Cambodian orphanage or residential care institution? Yes No Choose not to respond Question 36 Why or why not? Question 37 What do you think would be the best way to get a message out to potential donors/volunteers to provide more information about orphanages or residential care institutions? Please provide up to three ranked options by dragging the message mode (left column) to the ranking option (right column) 54 55 ANNEX II: QUALITATIVE DATA CODEBOOK Name Description* Additional services need Services that respondents identify that they need or would like to have in their community that they do not currently have access to CCWC member Child characteristics What children are like, attributes, behavior Childcare Responses relating to childcare tasks, duties, responsibilities, challenges Community members Community support Debt Decision making Responses about decision making within families, including gender disparities, who makes decisions about what Donations Current donations Donation history Drug or alcohol abuse Education References to education, studying, importance of, etc. . Family care Descriptions of children being cared for in family, positive and negatives Family, family structure Descriptions of family composition, household members, which family members live together Grandparents Food access Gender Husband, male, father primary Men on own or in role as husband, or father are described as primary or primarily involved in activity like childcare or decision making Together Descriptions of cooperative actions like making decision making together (like husbands and wives), discussions, etc. Wife, mother, female primary Women on their own or in role as mother or wife are characterized as primarily responsible, primarily involved in activity like childcare or decision making Good intensions Description of meaning well, having good intensions, trying their best, doing best, regardless of outcome Government Government functions, services, or leaders excluding village chiefs Happy References to happiness, what makes people happy Health services Access to health services, either a clinic or specific health services Health status References to health status like being ill or having good health Hopes for children Responses about hopes, goals for children, what future people want, what they want children to achieve including future occupations 56 Impression of organization For donors, impression of the organization they were or are still involved with Improvement Responses about seeing improvement in RCI, in organization and their approach Infrastructure Initial involvement How donors initially became involved in organizations Knowledge about development and organizations Donors knowledge about development and organizations like what to look out for, what kinds of organizations to be wary of, etc., knowledge and understanding of how money is being used and what organizations are really doing Livelihoods Broader category about livelihoods like access to physical capital, livestock Agriculture Child labor Responses about what respondents specifically label as child labor Debt Responses about being in debt, borrowing money, interest rates, repaying loans Jobs Migration Money Access to money or cash specifically, positive or negative Poverty, not enough money References to poverty and being poor; references to lacking money or not having enough money Selling labor Media use, access Access to and use of media Message options For other community options described besides the ones we prompted for 1 - Messages on keeping families together As stated in focus group guide 2 - Messages on negatives when families split As stated in focus group guide 3 - Messages on education opportunities As stated in focus group guide 4 - Messages on negatives of RCIs As stated in focus group guide Donor messages Messages that respondents said would resonate with potential donors or volunteers Mode of communication Modes of communication to best reach community members or potential donors, including venue or group through which to reach people or incentive to encourage them to attend/engage Motivations to donate, volunteer Why donors donate and volunteer, why they think others do Motivations to continue to donate, volunteer Why they continued to donate/volunteer after organization changed model Reasons stopped donating, volunteering Why they stopped donating/volunteering with organization that they were involved with Recommend to friends Reasons why they would or would not recommend their friends donate or volunteer with RCI Negative Explicitly negative explanations, harms of something, negative attributes NGO, organization National or international NGOs 57 Orphan Children described as orphan, children whose parents have died Parents Positive Explicit positive explanations, benefits of something, positive attributes Problems, constraints Problems or constraints facing families, communities, parents, children Providing for children How families try to provide for their children economically and otherwise, what is involved or needed to care for them RCI Future or fate of children in RCI Perceptions of what does or will happen to children in RCI, what has happened to children in RCIs that they know, what they have heard has happened Impression of RCIs Community, donor, volunteer, KIIs impressions and perceptions of RCIs including role in community, function, positive or negative Placing children in RCIs Information about how people place children in RCIs, about how RCIs try to recruit children, the reasons why people place children in RCIs, who helps them and makes decision about it RCI, in community Respondent identifies RCI as in community, including information about that RCI RCI, none in community Respondent identifies that RCI is not located in community, may be nearby or still have effect on them Religion Mention of religious affiliation (excluding donor demographic question) Buddhism Service access and sources Sources of services mentioned; access to services, positive or negative Sources of information Sources of information that respondents recommend to spread important messages Teacher Travel and tourism Traveling to Cambodia for work, holiday, or volunteering; tourism and perceptions of it and the tourism industry Not visited Cambodia Donor has not been to Cambodia Visited Cambodia Donor has been to Cambodia, regardless of trip purpose Types of communication or information sources Advertisement Announcement Email communication Information about email communication from donors Facebook Information board In-person Internet Letter Meeting 58 Mobile phone Newspaper Other social media Radio TV WhatsApp Village or commune chief or leader Role of village/commune chief or leaders, importance of, as source of help or source of information Village volunteer Violence References to violence including child abuse and sexual violence Water Access to water and rainfall Worry Descriptions of experiencing worry *Note: Descriptions are not included for clear, discrete terms like teacher and Facebook. 59 ANNEX III: BIBLIOGRAPHY BCN. “Collected Viewpoints on International Volunteering in Residential Care Centres: An Overview.” Better Care Network. 2014. Berens, Anne E. and Charles A. Nelson. “The Science of Early Adversity: Is there a Role for Large Institutions in the Care of Vulnerable Children?” The Lancet, vol. 386, no. 9991, 2015, pp. 388-398. Brown, Sally. “Travelling with a Purpose: Understanding the Motives and Benefits of Volunteer Vacationers.” Current Issues in Tourism, vol. 8, no. 6, 2005, pp. 479-496. Browne, Kevin. The Risk of Harm to Young Children in Institutional Care. Save the Children, 2009. Butcher, Jim and Peter Smith. “‘Making a Difference’: Volunteer Tourism and Development.” Tourism Recreation Research, vol. 35, no. 1, 2010, pp. 27-36. EMC (Emerging Markets Consulting), UNICEF. “Study on Funding of Residential Care Institutions in Cambodia” (forthcoming). 2016. Farley, James. “Potential Short-Term International volunteers’ perceptions of children’s Residential Care.” 2015. Guiney, Tess and Mary Mostafanezhad. “The Political Economy of Orphanage Tourism in Cambodia.” Tourist Studies, 2014, pp. 1-24. Gurman, Tilly A., Sara E. Rubin, and Amira A. Roess. “Effectiveness of mHealth Behavior Change Communication Interventions in Developing Countries: A Systematic Review of the Literature.” Journal of Health Communication, 17:82–104, 2012, pp. 82-104. Guttentag, Daniel A. “The Possible Negative Impacts of Volunteer Tourism.” International Journal of Tourism Research, vol. 11, no. 6, 2009, pp. 537-551. Indochina Research Ltd. “Report Presentation: Adolescent Engagement.” 2016. Jordanwood, M. “Protecting Cambodia’s Children? The Role of Commune Committees for Women and Children and Informal Community-based Child Protection Mechanisms in Cambodia”. World Vision Cambodia. 2016. Lumos Foundation. “In our Lifetime: How Donors can End the Institutionalisation of Children.” 2015. McGehee, Nancy G. “Oppression, Emancipation, and Volunteer Tourism: Research Propositions.” Annals of Tourism Research, vol. 39, no. 1, 2012, pp. 84-107. McGehee, Nancy G. and Kathleen Andereck. “Volunteer Tourism and the “voluntoured”: The Case of Tijuana, Mexico.” Journal of Sustainable Tourism, vol. 17, no. 1, 2009, pp. 39-51. 60 Ministry of Social Affairs Veterans and Youth Rehabilitation. “Mapping of Residential Care Institutions.” Preliminary Data Compilation and Findings. Phnom Penh: March 2016. Mostafanezhad, Mary. “‘Getting in Touch with Your Inner Angelina’: Celebrity Humanitarianism and the Cultural Politics of Gendered Generosity in Volunteer Tourism.” Third World Quarterly, vol. 34, no. 3, 2013, pp. 485-499. Nelson, C. A.,3rd, C. H. Zeanah, N. A. Fox, P. J. Marshall, A. T. Smyke, and D. Guthrie. “Cognitive Recovery in Socially Deprived Young Children: The Bucharest Early Intervention Project.” Science (New York, N.Y.), vol. 318, no. 5858, 2007, pp. 1937-1940. Proyrungroj, Raweewan. “Orphan Volunteer Tourism in Thailand: Volunteer Tourists' Motivations and on-Site Experiences.” Journal of Hospitality & Tourism Research, 2014, pp. 1-24. Reas, P. J. “Boy, have we Got a Vacation for You”: Orphanage Tourism in Cambodia and the Commodification and Objectification of the Orphaned Child.” Thammasat Review, vol. 16, no. 1, 2013, pp. 121-139. Smith, Chris, Uk Vannak, Ly Sokhey, Thoai D. Ngo, and Judy Gold. “Mobile Technology for Improved Family Planning (MOTIF): The Development of a Mobile Phone-Based (mHealth) Intervention to Support Post-Abortion Family Planning (PAFP) in Cambodia.” Reproductive Health, vol. 13, no. 1, 2016, pp. 1. Smyke, Anna T., Sebastian F. Koga, Dana E. Johnson, Nathan A. Fox, Peter J. Marshall, Charles A. Nelson, and Charles H. Zeanah. “The Caregiving Context in institution‐reared and family‐reared Infants and Toddlers in Romania.” Journal of Child Psychology and Psychiatry, vol. 48, no. 2, 2007, pp. 210-218. UNICEF. “Residential Care in Cambodia.” Accessed September 23, 2016. UNICEF. “With the Best Intentions.” A Study of Attitudes Towards Residential Care in Cambodia. Phnom Penh, 2011. van Doore, Kathryn, Laura Healy, and Megan Jones. “Mapping Australia’s Support for the Institutionalisation of Children Overseas.” 2016. Van Olmen, J., G. M. Ku, M. Van Pelt, J. C. Kalobu, H. Hen, C. Darras, K. Van Acker, B. Villaraza, F. Schellevis, and G. Kegels. “The Effectiveness of Text Messages Support for Diabetes Self-Management: Protocol of the TEXT4DSM Study in the Democratic Republic of Congo, Cambodia and the Philippines.” BMC Public Health, vol. 13, no. 1, 2013. Wearing, Stephen. Volunteer Tourism: Experience that make a Difference. CABI, 2011. 61 ANNEX IV: RF MERL OVERVIEW Rapid Feedback Monitoring Evaluation Research & Learning (Rapid Feedback MERL) Rapid Feedback MERL (RF MERL) brings learning to the early stages of an activity in a rigorous, empowering, and flexible way that works to inform implementation. We use a variety of monitoring, evaluation, and learning methodologies in collaboration with USAID and implementing partners to optimize design and implementation decisions, and continually adapt to get to better activities faster. The Challenge: Few activities build experimentation and feedback into implementation in ways that work to inform decision-making and lead to learning and adaptation within the activity cycle. Even those with a strong evidence-based design often implement activities using assumptions about their efficacy that could benefit from rigorous testing. Furthermore, implementers may not use monitoring systems to gather evidence about what is working well in the activity and what is not in order to help improve ongoing implementation. And, while independent impact evaluations may provide important evidence, that evidence often comes too late to improve the activities assessed before the end of the activity cycle. What is Innovative about RF MERL? The RF MERL Consortium applies proven evaluation methods to test the effectiveness of specific components of an activity against alternative intervention options. The approach involves rigorously testing the success of two or more alternative intervention options at achieving short-term outcomes. This is done in rapid cycles to allow for timely feedback and course adjustment earlier than is typically done using standard methods. RF MERL evaluations may include use of advanced statistical techniques to improve statistical power and reduce sample size requirements, as well as the use of tools to achieve rapid data collection. These tools might include cell phones and tablets for survey implementation, SMS and interactive voice response (IVR) technology for remote data collection, and geospatial imagery from satellites. The RF MERL Approach: RF MERL first works to understand the problem that the activity aims to address. Through formative research, we identify two or more alternative intervention options to test, and the short-term outcomes that will be used to determine success. The implementers and RF MERL Consortium then conduct “Feedback Experiments.” Feedback Experiments encompass the wide range of testing approaches that are at the heart of RF MERL, with a focus on testing multiple intervention options to identify the optimal choice. They can range from small-scale testing of program components (e.g., using message A vs. message B to increase participation) to testing alternative interventions (e.g., encouraging reading to children via pre-natal education vs. a community activist approach). A variety of methods can be used, and are designed to be as rigorous as possible to ensure that results can be used to inform decision-making about activity design and implementation. RF MERL is tailored to our partners’ learning needs, the local context, and the realistic trade-offs between rigor, time, and budget. This mechanism is particularly useful at the early stages of a project when there is uncertainty about the specific intervention options. RF MERL is appropriate when used alongside an activity that is being tailored or adapted to a new setting, moving from a pilot to scale‐up phase, or considering alternative strategies to promote greater engagement or participation of potential beneficiaries. RF MERL can also be used when project teams are interested in "unpacking" programs to identify the most effective elements or understanding why a specific activity has not yet achieved its intended results. 62 U.S. Agency for International Development 1300 Pennsylvania Avenue, NW Washington, DC 20523