In many crucial ways, the world is a better place for children now than it was just a few decades ago. This is due in large part to the efforts of donors, NGOs, multilaterals, governments, and individuals who were determined to make a difference. One would think that the public would be celebrating these unprecedented gains while enthusiastically redoubling their support of efforts to eliminate child hunger. Instead, almost half (47 percent) of Americans believe that more spending from the US and other wealthier countries will not lead to meaningful progress in improving health for people in developing countries.[1]Some are losing hope that they can make a difference. Why is this the case? And what can we do about it?
It has been comforting and liberating for me, over the years, to better understand how our nature as humans is to often see things from a smaller, more provincial frame, and to better understand how our tendency to forgetfulness can lead to the wrong conclusions. It is liberating because – knowing our natural tendency – I am more likely to shift my frame of reference, step back, and question whether the world situation is as dark, threatening, and disparaging as we sometimes are led to believe. For example, teen births and pregnancies in the US have plummeted over the past two decades, decreasing by 50% between 1991 and 2010.[2] Despite this, a 2013 survey of adults aged 18 and older by the National Campaign to Prevent Teen and Unplanned Pregnancy[3] found that 49% of Americans thought the teen pregnancy rate had increased over the past two decades and 25% believed that it had stayed about the same.
This tendency is due in part to what psychologists and behavioral economists call the “availability heuristic,” which says that when an infrequent event can be brought to mind easily and vividly, we often overestimate its likelihood.[4] Our sensationalistic 24-hour news cycle does a good job of making troubling events both easier to recall and vivid. Hearing about two well-publicized murders in the city in which we live can make us feel like crime is on the rise, despite the fact that good statistics may show a substantial decrease in crime. In fact, since 2001, a growing percentage of Americans believe that crime is on the rise when the rate has actually fallen substantially.[5] Ironically, while television shows like 16 and Pregnant can make a problem like teen pregnancy more easy to bring to mind (and hence to overestimate), their effect on social norms can help decrease the problem. In fact, a study by Melissa Kearney and Phillip Levine[6] found that 16 and Pregnant “ultimately led to a 5.7 percent reduction in teen births in the 18 months following its introduction” and accounted for “around one-third of the overall decline in teen births” over that period.
In a similar way, seeing pictures of malnourished and dying children on television can give the impression that these problems are on the rise despite the progress we have made. As Chief Program Officer for Feed the Children, I imagine that I have contributed to that problem in some way, since it is common among development organizations to focus mostly on the problem – dying and malnourished children, women dying in childbirth – as a way to engage with people in the Global North and to encourage them to take action through advocacy and donating funds for development projects. But the truth is that – on many fronts – things are better now for children than they have ever been in the past. More children are now surviving infancy. Despite population growth, over the past two decades, the number of child deaths in the world has been cut in half. More women are surviving childbirth; the proportion of women who die during or around childbirth dropped by 45 percent over the past 14 years. People also have more income; in 20 years, we have cut in half the absolute poverty rate, the proportion of people living on $1.25/day or less. There are also fewer new cases of HIV; the number of new HIV infections per 100 adults declined by 44 percent over about a decade. Yes, there’s still a lot to do, particularly in nutrition; nearly one in eight people in the world are still chronically malnourished. But we need to celebrate these improvements and realize that we have realized a lot of gains. Along with some of the tragedies we have seen, there are a lot of “bright spots” – and more light than darkness.
There are two main problems with our usual tendency to “focus on problems” and to underestimate gains that I have seen, both at the community level in the 25 countries in which I have worked and at the donor level. I discovered one of those problems during my training on Motivational Interviewing (“MI”).[7] MI is a psychological and behavior promotion approach that has been used successfully with some of the most intractable problems that individuals face (e.g. alcoholism, drug abuse, binge eating). In MI, an individual is often assessed early on for howimportant they think it is to overcome the problem they are facing, and their degree of perceived self-efficacy (i.e., how strongly one believes in one’s own ability to overcome the problem). However, when both importance and self-efficacy are low, the practitioner is urged to focus on perceived self-efficacy first. The reason for this is that when you feel trapped in a situation with little in the way of skills to overcome it, it does little good to have someone beating you over the head with how important it is that you change your behavior. (This is reminiscent of the early findings on “learned helplessness” by Martin Seligman.) It can create despair and cognitive dissonance: “I must change … but I cannotchange.”
By using techniques to “roll with [the client’s] resistance,” while slowly helping the person to build hope and skills for change, practitioners have been able to help many people to make astounding changes in their lives. I believe that our overemphasis on the importance of the problems in the world – rather than our ability to change them, and the positive changes that are occurring – has led to some learned helplessness and despair in the donor community (including the American public) and a false worldview concerning the impact we are making in improving life for children. It is stealing our enthusiasm and creating cynicism.
A focus on problems also can lead us to develop a worldview that increasingly leads to our ignoring information which conflicts with our beliefs through “confirmation bias,” where we search for and interpret information in a way that confirms our existing beliefs or hypotheses. If we believe that the world is getting worse for children, it becomes harder to believe the success stories and breakthroughs and we run the risk of selectively looking for – and exclusively talking about – the existing challenges and tragedies. But is it possible that this focus has led to the sad situation where the only option on which Americans can agree for cutting government spending (of 19 options presented by pollsters in a 2013 poll[8]) is foreign aid? If it appears that the world is a darker and darker place despite the billions spent, why would they want to sink more money into foreign aid? The fact that the average American believes that 25 percent of the US federal budget goes to foreign aid, despite it being less than one percent, makes it even more clear that we have failed in telling the story of foreign aid in a way that motivates people to further engagement.
Another downside of focusing on problems is that it can lead to an ignorance of the coping mechanisms that many people – and organizations and governments – already have in place that can to be brought to scale to fight poverty. According to UNICEF’s State of the World’s Children, 35 percent of Sudanese children were moderately or severely stunted between 2008 and 2012. This is tragic … but what about the other 65 percent? Given the economic and health conditions there, how is it possible that parents of 65 percent of the children – many of them still desperately poor – have children that are growing normally? These are the types of questions that we need to be asking and researching more often.
Many people are under the impression that ideas, interventions, programs, and coping mechanisms that are effective will naturally spread, and those that are not effective will naturally die out as more and more people notice the good or the bad outcomes associated with a program or an intervention. This is not always the case. Sometimes word of mouth or anecdotal evidence from individual successes can paint an image that does not hold true in the aggregate. Timothy Wilson, a professor at the University of Virginia, details several examples of this in his book Redirect. One is the Drug Abuse Resistance Education (“DARE”) program. DARE is a school-based program that reaches 36 million children worldwide in 43 countries. Its purpose is to combat the use of alcohol, drugs, and tobacco and it has rapidly spread across several countries. The problem is that it does not work. For example, in 1992, researchers at Indiana University found that those who completed DARE subsequently had significantly higher rates of hallucinogenic drug use than those not exposed to the program. A 1998 study by the University of Maryland, funded by the National Institute for Justice, concluded, "DARE does not work to reduce substance use."[9] A ten-year study by the American Psychological Association released in 2006 involving one thousand DARE graduates found “no measurable effects” of the program.
Another example is the Scared Straight program where kids who are at risk (e.g. arrested once for a petty crime) visit jails where prisoners tell them how hard jail life is, yell at them, and generally scare them about prison. Randomized studies of Scared Straight programs found that – not only do they not reduce the likelihood that kids will commit crimes – they actually increase criminal activity in children who go through the program. This was the same finding in every single study – the increase ranged from one percent to 30 percent, with an average of 13 percent more crime. Despite these findings, it is easy enough to find people who will give excellent individual testimonials as to how the program helped them, or helped the children in their community.
So if we cannot rely on positive anecdotes on programs and not every organization or community has the resources to conduct their own highly rigorous research, what tools are available to draw out what actually works to help children grow and thrive given a particular place or situation? It can be useful to think about approaches and tools for two main purposes: (1) finding out which practices/behaviors (of parents in particular, but also of health workers and governments) lead to better survival and development of children; and (2) finding out why some people adopt those behaviors/practices while others do not (be they parents or policy makers – but we will focus on parents).
We have a rich body of scientific evidence about interventions to promote practices that parents can adopt to improve child health and nutritional status in general. However, many of the interventions result in improved outcomes in some communities, but not in others. By using formative research to better understand the people who they wish to serve, organizations and governments can create programs that are specific to the needs of local populations, help ensure that programs are feasible and acceptable to community members, and improve relationships between organizations, government agencies, and communities.
Formative research can also help practitioners identify priority practices to promote, since there are only so many hours in a day, and it is often difficult to find the “contact time” with parents needed to make changes happen in all of the behaviors that we know may improve child growth and health. It can help development workers to prioritize, and focus on promoting those behaviors that can make the largest difference in child growth in a given setting rather than taking a “shotgun approach” that targets a long list of factors that agencies think are important. One formative research methodology that has been shown to be very helpful is the Positive Deviance approach[10] that focuses on reducing malnutrition, now responsible for 45 percent of all child deaths.[11] Interventions that use a Positive Deviance approach identify the successful practices and strategies that low-income parents with normally developing children are using and then work to scale up those successful behaviors. A key tool in that approach is a formative research tool called the Positive Deviance Inquiry (“PDI”). The PDI is basically a case-control type survey where the practices of low-income parents of children one to five years of age with a good weight-for-age are compared to the practices of low-income parents of children one to five years with a poor weight-for-age – those that are malnourished – in order to identify behaviors which are most highly associated with proper child growth.
Some programs have done this inquiry in a simpler and more participatory way, using local mothers to carry out the inquiry, often without written questionnaires. With this approach, local women are taught to talk with the mothers of children with good nutritional status, asking them questions to discover how they care for their children in three main areas: food and feeding practices, health-seeking practices and home care of sick children, and affection and caring for the child. This type of inquiry is good at getting mothers engaged with the process and connected to the “positive deviant” mothers from whom they can learn. The findings are used to prioritize the practices that will be promoted with mothers of malnourished children who come together for a ten-day rehabilitation process based in mothers’ homes (around the family “hearth”). This is the standard PDI used in the highly successful Positive Deviant/Hearth Nutritional Rehabilitation model.
Another more rigorous form of PDI was developed by practitioners who currently work with Feed the Children (Phil Moses and myself, during our work with Food for the Hungry), and is called the Local Determinants of Malnutrition (“LDM”) Study. The LDM Study uses individual interviews conducted by trained project staff and a spreadsheet or statistical software to identify statistically significant differences between the key practices of the two groups – mothers of children who are growing well and mothers of children who are not. This type of inquiry can identify smaller differences in parent practices that may be harder to identify in traditional PD/Hearth programs. The LDM Study also includes questions about potential causes of malnutrition that may be found in some project areas, but not others, such as alcoholism among family members, depression in the mother, maternal diet during breastfeeding, and intake of specific nutrients by the child (e.g., magnesium, potassium, and phosphorus). Results of the study are then used to decide which practices should be promoted at the household level, and which environmental factors the project should try to change to bring about improvements in nutritional status.
Both the PDI and the LDM studies can be conducted rapidly (within one to two weeks), and both have been successful in identifying local practices of parents that can make a big difference in their children’s growth and survival. For example, using the traditional PDI that is part of a PD/Hearth Nutritional Rehabilitation program, Save the Children (“SC”) project staff and participants in Vietnam found that low-income mothers who had children that were growing well took advantage of tiny shrimps, crabs or snails, and greens that grew abundantly in the rice paddies that other mothers did not consider a food source appropriate for a child. SC brought an estimated 50,000 children out of malnutrition between 1991 and 1999 in Vietnam using this approach. More importantly, years after the project, the younger siblings of these children – many of who were not yet born at the time that the nutrition program was implemented – were also better nourished. The proportion of children in the program who were severely malnourished decreased by 74 percent, from 23 percent to 6 percent (p<0.001).
LDM Studies have now been completed in at least seven countries (most by Food for the Hungry). Those studies were successful in identifying differences in care provided by parents. Here are some examples of some of the practices (behaviors) and conditions that were found by Food for the Hungry to be associated with good child growth using the LDM Study:
- In a 2013 South Sudan LDM study, the average depression score of mothers was better amongst mothers of positive deviant (“PD,” in this case meaning properly growing) children when compared to mothers of children who were not growing well (OR=1.1852, p=0.009) even after controlling for the child’s age, mother’s age, and marital status.[12] We are just beginning to see the importance of maternal depression in child growth. A meta-analysis by Pamela Surkan published recently indicated that elimination of maternal depression could possibly reduce child malnutrition by 23-29 percent.[13] An earlier randomized controlled trial in Uganda proved that depression could be dramatically reduced in four months through a simple community-level intervention called Interpersonal Therapy for Groups.[14]
- In a Mozambique LDM study, mothers of positive deviant children were seven times more likely to say that they usually or always completely emptied their breasts when breastfeeding their child than mothers of malnourished children (OR=7.09, p=0.006).[15] This is a good example of a practice that is easy to adopt, but where mothers may be uninformed about its benefits. By completely emptying the breast (before switching to the other breast to feed), mothers are more likely to give their child more of the higher-fat and higher-protein hind milk.
- In a 2008 Ethiopia LDM study, 35 percent of the PD children vs. eight percent of the malnourished children defecated in a proper spot the last time they defecated (p=0.001).[16] Mothers of PD children in this study also had an average hygiene practices index score of 2.61 vs. 2.06 for mothers of malnourished children (p=0.03). (This was also found in LDM studies in Burundi and Kenya.) Since that time, it has become clear that water, sanitation, and hygiene practices (such as defecation in the open) are more highly associated with stunting than we previously thought, perhaps because of their link with a common condition called environmental enteropathy.[17]
- In a Kenya LDM study, mothers of positive deviant children were found to be away from their child an average of 5.0 hours a day vs. 6.7 hours a day for mothers of malnourished children.[18]
- In an LDM study conducted in the Democratic Republic of Congo, mothers of malnourished children were 4.5 times more likely to have introduced semi-solid/mashed foods late (at nine months of age or after) as compared to mothers of positive deviant children.[19]
One striking finding that can be seen looking across these studies is that many of the associations found in one project location and country were quite different from those found in another project location and country. Practitioners could just try to promote the dozens of helpful behaviors everywhere that children are malnourished. But given time and resource constraints, it makes a lot more sense for practitioners to not rely on a “shotgun approach” and instead do these sorts of rapid studies in each project location in order to identify priority behaviors that can make the most difference.
As I have used these methods to identify key positive deviant behaviors, another advantage emerged: they bring hope. They help both project staff and local people to better understand that there are things that parents can do now with their current resources to lower child malnutrition and the child deaths associated with malnutrition. They remind community members of their current strengths and resilience, and that any parent can be heroic and help their children to survive and thrive despite their impoverished situation and all of the challenges of living where they live. Notwithstanding these positive results and replications in other settings, there are – to my knowledge – no national Ministries of Health aside from Vietnam’s and Burundi’s that have adopted the PD/Hearth model or the use of Local Determinant of Malnutrition Studies (or other PDIs) as a standard methodology to combat malnutrition nationally.
Identifying positive deviant, adaptive behaviors that parents use that lead to better survival and development of children is only half the battle, though. It is also important to find out why some people adopt those behaviors/practices while others do not. History is littered with good ideas and useful technologies that had plenty of science behind them, but were never adopted by large numbers of people. About half of child deaths could be prevented by adoption of a relatively small set of practices if parents adopted them universally.
For example, 13 percent of all child deaths could be averted if mothers followed recommendations regarding breastfeeding – immediate, exclusive, and continued – and 15 percent of child deaths could be averted if all parents used oral rehydration serum (“ORS”) at home when a child has diarrhea.[20] ORS was first used at Bangladesh’s Cholera Research Laboratory (now the International Centre for Diarrhoeal Disease Research, Bangladesh, “ICDDR,B”) 35 years ago. The introduction of ORS and the uptake of behaviors that can decrease diarrheal deaths (e.g. continued breastfeeding and complementary foods) led to dramatic declines in infant and child deaths worldwide, with 4.6 million diarrheal deaths per year prior to 1980 yet only 2.6 million diarrheal deaths per year between 1990 and 2000.[21] However, many countries still do not have particularly high ORS adoption rates. Twenty developing countries have entirely failed in promoting good diarrhea management, with less than 25 percent of infant and child diarrhea episodes treated with ORS.[22] While there are many causes for this failure (including poor supply chain management), failure to use adequate and appropriate behavior changes methods are undoubtedly a big part of the problem.
Identifying the reasons for these failures in promoting behavior change requires tools that many practitioners still fail to use. Since the barriers and enablers of these lifesaving practices can vary from project area to project area, easy-to-use tools are needed to identify the behavioral determinants of these practices in eachproject area. Two tools for this purpose that are gaining popularity amongst community health practitioners (especially among US NGOs) are Barrier Analysis (“BA”) and Doer/Non-Doer Analysis.[23] Both of these tools have a similar focus on what is going right, on a community’s positive assets, and – in this case – a focus on the people who have already adopted the behavior. A survey is conducted of parents (often mothers) who have adopted a childcare behavior (e.g. use of ORS, hand washing with soap, water purification) and their responses are compared to parents who have not adopted the behavior in the same geographical area. The questions focus on 12 important behavioral determinants that often influence behavior (e.g. perceived self-efficacy, perceived social norms, perceived positive and negative consequences). The results of this rapid study can then be used to adapt project messages and curricula to focus on the key behavioral determinants that are found.
As practitioners have used Barrier Analysis, in addition to increased uptake of important behaviors, another advantage has emerged: it can humanize us and help us to see things from community members’ points of view. In a Catholic Relief Services natural resources management project in Guatemala, staff members were trained in Barrier Analysis as part of an overall Designing for Behavior Change workshop.[24] In addition to seeing large changes in behavior in the areas where they used Barrier Analysis (in comparison to decreases in the behavior in nearby project areas where CRS did not conduct the study), they also mentioned changes in the thinking of their staff members. When asked about the helpfulness of developing a behavior-change strategy based on the results of the BA study, one staff member commented, “Yes, this completely changed our way of thinking. We no longer think in terms of ‘we’ and ‘them’; rather we are a team with the participants in finding solutions to the barriers.” Another staff member said, “It never occurred to us before [the training] to figure out the barriers or what makes people want to change. We wasted so much time and energy repeating the benefits over and over, then feeling frustrated because no one adopted the new practices.”[25]
This scenario is playing out in many places in the world as practitioners begin to use more up-to-date methods to help people change, including formative research. It’s time for donors to increasingly demand that local, rapid formative research be a part of more and more development projects. We all need to see things from the eyes of those who are most in need, and to value what they are already doing that is successful – and formative research tools can help us to do that.
It will take better-informed villages for practitioners to continue to make progress against child deaths, malnutrition, and other development problems. We need villages that understand themselves – not just their problems, but their resources, their successful coping strategies, and why some of their low-income members are making progress and have healthier children despite their life conditions. It is urgent that we refocus on these community assets and scale up the use of formative research tools that help communities to “find their bright spots.”
[1] Henry J Kaiser Family Foundation (2012). US Global Health Policy: 2012 Survey of Americans on the US Role in Global Health. Seehttp://kaiserfamilyfoundation.files.wordpress.com/2013/07/8304_2012-global-health-findings-final.pdf.
[2] Kost K and Henshaw S. (20xx). U.S. Teenage Pregnancies, Births and Abortions, 2010: National and State Trends by Age, Race and Ethnicity. Guttmacher Institute. Available here: http://www.guttmacher.org/pubs/USTPtrends10.pdf
[3] Poll conducted in December 2013 by The National Campaign by Social Science Research Solutions, an independent research company. The margin of error was +/-3.1% at the 95% confidence level. Fact sheet available here:https://thenationalcampaign.org/sites/default/files/resource-primary-download/december.pdf
[4] Tversky, A; Kahneman (1973). "Availability: A heuristic for judging frequency and probability". Cognitive Psychology 5 (1): 207–233. doi:10.1016/0010-0285(73)90033-9
[6] Kearney M and Levine P. (2014) Media Influences on Social Outcomes: The Impact of MTV’s 16 and Pregnant on Teen Childbearing. Presentation at the Apsen Institute, 9 June 2014, Washington DC. Paper available here: http://www.wellesley.edu/sites/default/files/assets/kearney-levine-16p-nber_submit.pdf
[7] See http://www.motivationalinterview.org for more details.
[8] See http://www.people-press.org/2013/02/22/as-sequester-deadline-looms-little-support-for-cutting-most-programs/
[9] Lawrence W. Sherman, Denise Gottfredson, Doris MacKenzie, John Eck, Peter Reuter, and Shawn Bushway. Preventing Crime: What Works, What Doesn’t, What’s Promising. Report for the National Institute of Justice. Chapter 5. School-based Crime Prevention, 1998.
[10] For more on the Positive Deviant / Hearth nutritional rehabilitation model, see this publication: http://www.ncbi.nlm.nih.gov/pubmed/12503241; these resources: http://www.coregroup.org/our-technical-work/initiatives/diffusion-of-innovations/84; and Ramalingam, Ben (2013-09-28). Aid on the Edge of Chaos: Rethinking International Cooperation in a Complex World (pp. 272-298). Oxford University Press, USA.
[11] See the Lancet’s Maternal and Child Nutrition Series athttp://www.thelancet.com/series/maternal-and-child-nutrition
[12] Personal communication with Sarah Borger, Senior Director of Health Programs, Food for the Hungry, 13 September 2014.
[13] Surkan P, Kennedy C, Hurley K, and Black M. (2011) Maternal depression and early childhood growth in developing countries: Systematic review and meta-analysis. Bulletin of the World Health Organization 2011; 89:608-615 http://www.who.int/bulletin/volumes/89/8/11-088187/en/
[14] Bass J, Neugebauer R, Clougherty K, Verdeli H, Wickramaratne P, Ndogoni L, Speelman L, Weissman M, and Bolton P. (2006) Group interpersonal psychotherapy for depression in rural Uganda: 6-month outcomes
Randomised controlled trial. British Journal of Psychiatry, 188, pp. 567-573.http://www.aliveandthrive.org/sites/default/files/IPT-G%20Results,%206m%20Post-Project.pdf
Randomised controlled trial. British Journal of Psychiatry, 188, pp. 567-573.http://www.aliveandthrive.org/sites/default/files/IPT-G%20Results,%206m%20Post-Project.pdf
[15] Davis et al. (2010). Local Determinants of Malnutrition: An Expanded Positive Deviance Study. Food for the Hungry. Full study is available here:http://www.fsnnetwork.org/sites/default/files/an_expanded_positive_deviance_study.pdf
[16] Ibid.
[17] Lin A, Arnold B, Afreen S, Tarique R, Huda M, Haque R, Raqib R, Unicomb L, Ahmed T, Colford J, and Luby S. (2013). Household Environmental Conditions Are Associated with Enteropathy and Impaired Growth in Rural Bangladesh. Am J Trop Med Hyg 12-0629; Published online April 29, 2013.
[18] Ibid.
[19] Ibid.
[20] Jones G, Steketee R, Bhutta Z, Morris S. and the Bellagio Child Survival Study Group. "How many child deaths can we prevent this year?" Lancet 2003; 362: 65-71.
[21] Keusch GT, Fontaine O, Bhargava A, et al. Diarrheal Diseases. In: Jamison DT, Breman JG, Measham AR, et al., editors. Disease Control Priorities in Developing Countries. 2nd edition. Washington (DC): World Bank; 2006. Chapter 19. See http://www.ncbi.nlm.nih.gov/books/NBK11764/pdf/ch19.pdf
[22] Wilson, S; Morris, S; Gilbert, S; Mosites, E; Hackleman, R; Weum, K; Pintye, J; Manhart, L; Hawes, S. (2013) Scaling up access to oral rehydration solution for diarrhea: Learning from historical experience in low– and high–performing countries. J Glob Health. Jun 2013; 3(1): 010404. Available here:http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3700030/
[23] Davis Jr., Thomas P., (2004). Barrier Analysis Facilitator’s Guide: A Tool for Improving Behavior Change Communication in Child Survival and Community Development Programs, Washington, D.C.: Food for the Hungry. The original facilitator’s guide for Barrier Analysis is available here:http://www.coregroup.org/storage/Tools/Barrier_Analysis_2010.pdf. An updated manual on BA is available here:http://www.coregroup.org/storage/barrier/Practical_Guide_to_Conducting_a_Barrier_Analysis_Oct_2013.pdf
[24] Food Security and Nutrition Network Social and Behavioral Change Task Force. 2013. Designing for Behavior Change For Agriculture, Natural Resource Management, Health and Nutrition. Washington, DC: Technical and Operational Performance Support (TOPS) Program. The training manual for the Designing for Behavior Change workshop is available here:http://www.fsnnetwork.org/sites/default/files/combineddbc_curriculum_final.pdf
[25] Impact of DBC Training on Strategy on Behavior Change Uptake: CRS Guatemala Title II Program, Natural Resources Component. Unpublished project document received from final evaluation consultant and presented at the 2012 TOPS Capacity Strengthening Workshop in Addis Ababa, Ethiopia.
