SPRING/Digital Green Collaboration Progress Report

Filming the community video “A Good Start to Exclusive Breastfeeding” in the village of Halbawa Salifou in Guidan Roumdji, Maradi, Niger Photo credit: Peggy Koniz-Booher, SPRING

SPRING is a USAID-funded five-year project committed to reducing anemia and stunting among women and young children through the expansion of effective social and behavior change communication (SBCC) nutrition programs. SPRING is working to achieve this by identifying and testing proven or highly promising SBCC tools and approaches. Starting in 2012, SPRING and Digital Green (DG) began testing the feasibility of their innovative “human-mediated digital learning approach”—which previously focused on the promotion of improved agricultural behaviors—to promote high-impact maternal, infant, and young child nutrition (MIYCN) and hygiene practices in the Keonjhar District of Odisha, India. The results of the feasibility study, conducted by the International Food Policy Research Institute (IFPRI) and the London School of Tropical Medicine and Hygiene (LSTMH), were very supportive of scaling up the participatory model in India and in other contexts.¹

In 2015, SPRING and DG adapted the approach in partnership with three projects in the Maradi region of Niger: the Resilience and Economic Growth in the Sahel – Enhanced Resilience Program (REGIS-ER); Livelihoods, Agriculture and Health Interventions in Action (LAHIA); and Sawki, led respectively by the National Cooperative Business Association CLUSA International (NCBA CLUSA), Save the Children, and Mercy Corps. The objective of this one-year pilot project was to establish the impact and scalability of this collaborative model, and to document the process and costs associated with its implementation in the resilience context of the Sahel.

Intervention

Pilot Project Overview

With this pilot project in the Maradi region of Niger, SPRING and DG evaluated whether the community video approach that was successful in India is feasible and acceptable in the resilience context of Niger. SPRING and DG chose Niger to test the adaptation based on their mandate to assist the REGIS-ER resilience program by ensuring that MIYCN and hygiene SBCC are highly effective. SPRING and DG partnered with REGIS-ER, LAHIA, and Sawki to test the SPRING/DG collaborative approach in their respective intervention villages in Maradi. Implementation began in January 2015, with the identification of 20 villages in the Maradi region (table 1). In each village, SPRING and its partners worked with four previously established community groups. Membership in these groups ranged from 10–25 members. It was estimated that the project would reach approximately 1,500 community members, at least 70 percent of whom would be adolescent girls, pregnant women, and/or mothers of children under the age of two. Videos on high-impact MIYCN and hygiene practices were produced by a local video production community team and disseminated and discussed at community meetings.

Table 1. SPRING/DG Partners and Implementation Villages

Young girls taking care of their siblings, Baleyara, Niger, May 2014.

Project objectives—

1. Test the application of the SPRING/DG collaborative approach in the resilience context of Niger using videos developed by and for the community to promote high-impact MIYCN and hygiene behaviors.
2. Assess recall and comprehension to determine if women (adolescent, pregnant, lactating, and with children under two), and other caregivers and influencers (mothers-in-laws, husbands/fathers) who attended the community video screenings and facilitated discussions can recall and comprehend information presented in the videos.
3. Compare the actual adoption or promotion of behaviors to stated intention using data from three rounds of data collected at baseline; 1–2 months following the video screenings; and 2–4 months after exposure to measure intention to adopt or promote two specific priority behaviors (handwashing and responsive feeding).
4. Document the process capturing key elements of the pilot, including lessons from tool adaptation, implementation costs, tracking of behavior change, and acceptability of the model.
5. Assess the scalability and cost of the approach, and make recommendations about next steps.

Situational Analysis and Formative Research

SPRING initially conducted a desk review of published literature and unpublished program documents—including peer-reviewed and gray literature—related to nutrition and WASH practices in Niger, focusing on the region of Maradi. Based on the situational analysis, a number of information gaps were identified for further research on key MIYCN practices.

The purpose of the formative research was to address these gaps and explore the barriers and enablers of priority MIYCN practice uptake in relationship to other contextual realities (i.e., socio-cultural, economic). Results from the formative research clarified the relationship between target groups within the community and identified priority behaviors for video production.

Mahmud Hambali from Digital Green shows the process of storyboarding (with Save the Children SBCC advisor, Aboubacar Sale)

The research was conducted in a series of 10 focus group discussions (FGDs) and 12 in-depth interviews (IDIs) by two teams in two villages that represented the 15 villages in the commune of Guidan Roumdji, where the videos would be filmed. A total of 121 village members participated in the FGDs. Additionally, 12 IDIs with four women, four men, and four adolescent girls were conducted. The two mixed-gender research teams comprised program staff from REGIS-ER, Sawki, and LAHIA in Maradi.

Notable findings from the formative research:

• Handwashing is widely practiced for religious reasons but there are no fixed handwashing stations, nor do people typically use soap.
• Women receive no special care or food while pregnant or lactating, and multiple wives and favoritism are substantial barriers to women’s access to adequate nutrition.
• Complementary foods are not diverse and are too watery; young children between 6 and 24 months eat from the family bowl so the mother or other caregiver does not know how much an individual child has eaten.
• Breastmilk is often supplemented in the first six months and women stop breastfeeding when they become pregnant again.
• There is high interest and unmet need for family planning but lack of couple communication is a major barrier to FP use.
• Malaria and diarrhea are considered “normal” and many people do not know that these diseases are preventable.

The results of the formative research were fundamental to the adaptation of the SPRING/DG approach, informing the selection of priority nutrition behaviors for the ten videos, the design of a series of two-day nutrition sensitization trainings for key community agents, the overall strategy for video production and dissemination, and the verification points for the behavioral adoptions tracking plan.

Training

The project launched with a series of four different trainings. The first set of trainings, described below, targeted 40 community “mediators;” a man and a woman volunteer from each of the 20 villages. The supervisors of these mediators and the selected production team based in Guidan Roumdji also participated. Participants were divided into three groups, one in Aguié and two in Guidan Roumdji. Trainings were conducted entirely in Hausa over five days.

The first four-day training combined MIYCN/hygiene and dissemination, and was facilitated by national and international trainers. The first two days followed the MIYCN training manual that was developed by SPRING and adapted from the India community video project, the last two days focused on video dissemination During the second halve, received the standard DG video dissemination training, starting with how to use the equipment and ending with how to conduct home visits.

An additional five-day training facilitated by DG taught the three-person video production team and selected local partner staff to write storyboards and produce and edit videos. After that, DG gave a one-day training to staff on the routine program monitoring data system called Connect Online, Connect Offline (COCO), which is used to complement study findings with data on video production, dissemination, and adoption of practices.

The local video production team based in Guidan Roumdji, Maradi, Niger

Video Production

The SPRING team worked closely with DG and local partners to prioritize the MIYCN topics to be featured in the 10 videos. The teams collaborated in the development of a package of practices (POPs), which provided technical details for the production of the videos. POPs include the key practices to be captured in the videos, and the critical knowledge points that constitute “adoption” of the specific behaviors being promoted by the videos.

The local production team worked with SPRING, DG and partners to create storyboards, using the POPs as their reference. The storyboards were finalized and reviewed by SPRING and DG technical staff for accuracy. The videos were recorded in one day with inexpensive camcorders, external microphones, and tripods.

After filming, the field agent edited footage to 10-15 minutes. Each video was field-tested with target audiences to ensure that the content was clear and easily understood. When necessary, content was adjusted to accommodate feedback. Following final approval from the technical team, videos were uploaded onto low-cost mobile pico projectors for dissemination. The entire production process, from storyboarding through editing, took an average of five days per video.

Video Dissemination

The 40 volunteer community mediators disseminated the videos to village groups, including mother-to-mother support groups, écoles des maris (husband schools), adolescent “safe space” groups, and credit groups, every month. REGIS-ER and Sawki coordinated distribution in 15 selected villages in the commune of Guidan Roumdji, where the videos were also produced. LAHIA disseminated videos in five selected villages in the commune of Aguié.

Video being disseminated by the mediator in a women’s group in Maradi, Niger

In March 2015, the first video, on the importance of the first 1,000 days, was produced and disseminated to all 20 villages. SPRING launched the first six videos to the general public on its website and social media in September 2015. By the end of December 2015, 10 videos had been produced and disseminated.

The videos feature local women and men from their community demonstrating key behaviors and overcoming barriers discovered during formative research. Using local families allowed viewers to observe the practices in a familiar context, being conducted and promoted by people of similar means in their own language. By seeing a neighbor practice a behavior, viewers understood that they too had the ability and resources to do the same.

The videos were disseminated once a month to a group led by a mediator who was trained to pause the video at selected points to discuss questions, and to encourage adoption and/or promotion of the practices. This “human-mediated digital learning approach” uses dialogue and reflective processes among peers, instead of the traditional approach of outside experts telling people what to do.

To increase support for the behavior being promoted, each group member is asked to invite one other community member (e.g. spouse, mother, neighbor, leader) to come watch the video in that same week. These people are called the “influencers” and can be different each month, depending on the behavior being promoted.

While the videos are a great opportunity for discussion, it is the engagement and empowerment of people and social dynamics that make the model successful. Enthusiasm about new technology and innovation in general, combined with the thrill of appearing or seeing fellow community members on video, motivates individuals to participate. The power of positive identification with peers minimizes the distance between teacher and student, and maximizes adoption of the practices or behaviors being modeled.

Program Evaluation

The overall program evaluation used a mix-method approach that involved three rounds of quantitative surveys administered to program beneficiaries; routine monitoring through the COCO database; a costing analysis; and stakeholder workshops to gather qualitative data.

Research Questions

SPRING sought to answer the following evaluation questions to provide evidence of success for the proof-of-concept endeavor.

1. How did the SPRING/DG approach change knowledge, attitudes, intent, and adoption of handwashing and active feeding in the intervention target group? Did knowledge, intent, and adoption diminish or increase after initial trial of behaviors?
2. What was the cost per adoption and cost per person reached in the SPRING/DG project area implemented by the different partners? What is the cost of a potential scale-up?
3. How was the project perceived by the target population and the influencing groups? How, when, and why did the process of promotion and adoption take place? How did people perceive their “changed behavior”?

Methodology

The methods used to answer these research questions were—

1. Quantitative survey analysis of a cohort of women of reproductive age (WRA) over three time points.
2. Routine program monitoring of data from COCO to complement study findings with data on video production, dissemination, and adoption of practices.
3. Costing analysis of intervention in the two project areas.
4. Stakeholder workshop using FDGs on acceptability for different groups, including WRA, adolescents, men, as well as volunteers and supervisors implementing the program.

Quantitative Survey Analysis

The quantitative survey was conducted in all 20 villages; 15 REGIS-ER/ Sawki/LAHIA villages in the commune of Guidan Roumdji; and five Save the Children villages in Aguié. Data collection included all WRA with children ages 6–24 months who were members of the groups in the villages where DG videos were screened. In the REGIS-ER and Sawki villages, the study participants were primarily participants from (mother-to-mother) MTM/care unit women’s groups (ages 19–49), but included eligible adolescents 15–18-years-of-age from the two safe space (SS) groups. In the LAHIA villages, participating women aged 15–49 were beneficiaries from women’s groups and mixed gender groups.

Data collection occurred at three time points:

1. Baseline survey prior to screening of the first video
2. mid-line within two months of viewing each video
3. endline, two months after the videos were screened.

Routine Program Monitoring

Data, including the number of women present, video screened, questions asked, and participants’ stated intention to adopt or promote practices, were collected at each video screening. Stated intention to adopt or promote certain practices was verified by the mediators, who visited the participants in their homes. The definition of adoption for each video topic was based on critical points identified in the POP development process. The adoption verification process included assessment of participant knowledge, direct observation of uptake or promotion (if possible, given the particular behavior), and self-reported adoption. DG staff additionally conducted visits to 10% of households to cross-verify adoption findings. Results were reported on the COCO system, which was created by DG to be accessible by any web-enabled device to upload data in areas with limited Internet and electrical grid connectivity. Data were aggregated and analyzed on near-real-time analytics dashboards and mapped on the DG website. Additional data, such as demographics of attendees, attendance levels, and intention to adopt, was collected by dissemination facilitators and was made available outside the COCO dashboard.

Costing Analysis

Costing analyses calculate costs per adoption and per person reached, and present information on resources needed for future scale-up. A local consultant collected relevant data and the preliminary report for budgeting was shared with partners in October 2015. Clear tracking of the total number of participants in the intervention in both project areas was collected though COCO and served as the denominator to measure cost per exposure of the intervention. Cost data were compiled from each partner and analyzed alongside routine monitoring data on exposure from COCO. These costs were validated by cross-checking with the budget and by random audit. Cost effectiveness was calculated based on behavior change measured through the quantitative surveys at baseline and endline.

Participatory Stakeholder Workshop

From July 26–31, 2015 a participatory stakeholder workshop was held to assess the acceptability of the SPRING/DG pilot intervention in Maradi. The workshop was a collaborative effort between SPRING/DG; the Conception, Expertises, Suivi-évaluation, Appuis et Formation (CESAF); le Centre d’Études et d’Éxpérimentations Économiques et Sociales de l’Afrique de l’Ouest (CESAO); and Sahel Resilience Learning (SAREL). CESAO received funds from SAREL to support the participation of three experienced facilitators in the external evaluation. These facilitators led the workshops, made sure that all information was gathered during the group sessions, and conducted additional in-depth interviews with key participants. CESAF led the analysis of the results. The first part of the workshop solicited feedback from a representation of beneficiaries, and the second part engaged all community mediators and their supervisors.

Results

The main objective of this mixed-method evaluation was to measure the success of the SPRING/DG collaborative approach in the resiliency context of Niger. The results will inform future social and behavior change communication programming in Niger through REGIS-ER and other Food for Peace partners. Results were expected by December 2015 and will be released by May 2016.

Quantitative Survey: Baseline to Endline Key Behaviors

The quantitative evaluation was conducted to determine the impact of community video on two priority MIYCN and hygiene behaviors. The findings below are based on survey data collected at baseline (April 2015), a second survey implemented 2–8 weeks following exposure to the video, and an endline survey in August 2015. In table 2, results indicate that the presence of a handwashing station in the household increased from 14 percent at baseline to 48.8 percent in the second survey, and to 59 percent at endline. In addition, the percentage of households with a handwashing station with soap and water (assuming this equals routine and recent use), increased from 73.8 percent at baseline to 96.2 percent in August 2015.

Table 2. Handwashing Behavior

Results for the second behavior—active/responsive feeding and feeding from a separate dish—are presented in table 3. After watching the video, the percentage of women who said that they had a separate plate for their children increased from 69.8 to 96.6 percent. At baseline, 31.2 percent of women actively encouraged their child to eat. This increased to 70.2 percent after video exposure and 96.6 at endline. The active feeding scale was calculated based on 20 validated questions that measure the extent that a caregiver is actively feeding a child.

Table 3: Responsive Feeding and Feeding from a Separate Dish

COCO Adoption Results

Out of the 10 videos that were produced (Table 4), the data presented relate to the first three community videos that were produced and disseminated during the first three months of the project.

Table 4: Videos Produced and Disseminated (March–December 2015)

Dissemination Sessions for the Community

A total of 237 video dissemination sessions were held for the first three videos over a three-month period during four community group meetings in the 20 participating villages. This included 80 disseminations each for handwashing and active feeding videos and 77 disseminations for the first 1,000 days video.

Viewership and Adoption²

The first three videos reached 1,614 unique viewers (community group members³). On average, there were 16.6 unique viewers per dissemination session. Of all unique viewers, 1,578 (98 percent) adopted at least one of the three practices being promoted. The total number of adoptions for these three videos is 3,605,⁴ an average of 1,201 adoptions per video in the 20 villages.

Of the 20 participating villages, the highest numbers of “adoptions” were recorded in the village of Karazome, with a total of 254 adoptions. The lowest number was recorded in Katare Moussa, with a total of 90 adoptions. Both villages are located in the commune of Guidan Roumdji.

As shown in figure 1 below, the video with the highest adoption rate was on active feeding, which 95 percent of people adopted. Nearly 90 percent of people (1,172 of 1,323 viewers) adopted the 1,000 days information, which in this case meant that the viewer correctly explained the concept of the first 1000 days, for which no specific practice was promoted. For optimal handwashing, 63.4 percent of community group members (863 of 1,361 viewers) reported practicing optimal handwashing practices during follow-up visits.

Figure 1. Video: Viewers-to-Adoption Ratio

Videos were disseminated during community group meetings. Figure 2 shows the average adoptions per viewer in the three types of community groups with the greatest number of members and thus the greatest number of viewers. These were the safe space, écoles des maris, and care units for WRA (which include M2M and care) groups. Since the preliminary COCO report covered only the three videos that had been disseminated at that time, the maximum number of adoptions per group member was three.

The highest rate of adoption was achieved in the écoles des maris groups where the adoption rate was 2.7 per individual viewer. Among viewers from care units, the adoption rate was 2.4. There were 1.8 adoptions among adolescent girls who viewed the videos during safe space meetings.

These figures show a promising trend of adopting practices by members of targeted groups. For this preliminary analysis and in consideration of the small size of data, we calculated a ratio of total viewers in the group to total number of adoptions. In subsequent reports, we will calculate a ratio of unique adopters in the group to total adoptions, which will give us an exact ratio of the number of adopters to adoptions.

Figure 2. Groups: Viewers-to-Adoption Ratio

Costing Analysis Results

The costs associated with the project, estimated at $242,804 USD, were broken as described below.

• Two phases. A preparatory phase from August 2014 to December 2014, followed by an implementation phase from January to December 2015. From August 2014 to May 2015, the costs were collected from invoices, expense reports, and time sheets, while costs from June to December 2015 were estimated on the basis of the costs incurred in previous months.

Table 5. Breakdown of costs per phase

• Five processes. SP1 – strategic management of the project (identification of partners and beneficiaries, contracting, recruitment, planning, and budgeting); SP2 – adaptation of the DG approach to Niger (formative research, adaptation of manuals, and information system); SP3 - video production (training and production of 10 videos); SP4 – dissemination and outreach (training volunteers in MIYCN and hygiene practices, dissemination of videos, and home visits); SP5 – regional project coordination (organization and supervision of video production and dissemination).
• Four types of costs. 1) personnel costs, including the costs of international experts from the partners, not including staff costs at operational and managerial levels at headquarters; 2) travel expenses, including international travel; 3) the cost of equipment needed for video production and dissemination, and; 4) operational expenses for training venues and the regional office.

The breakdown of costs by process and type shows a particularly high cost for staff and travel (93 percent). These costs are largely related to support travel and staff costs for the DG and SPRING international experts, which is an anticipated initial investment for technical guidance that would presumably be scaled down as capacity is enhanced in subsequent scale-up phases.

Table 6. Breakdown of Costs by Process and Type

Two scenarios for scale-up were defined to assess the projected costs of expanding the project to other regions of Maradi and/or other areas of the Sahel.

• Scenario 1: Target 100 villages in Maradi region and disseminate the same 10 videos produced in the proof-of-concept project. Videos would not need to be adapted, given the similar context.
• Scenario 2: Target 100 villages in another region of the Sahel. This would involve formative research, adapting training tools, and producing new videos with new content.

To date, the cost-effectiveness of the project and the two scale-up scenarios are as shown in table 7.

Table 7. Comparing Cost-effectiveness of the Project and two Scale-up Scenarios

(*) Per video viewing

This study shows that the cost per video viewing in the proof of concept is relatively inexpensive and decreases exponentially as the project expands and takes advantage of economies of scale. Additionally, the proof of concept included substantial costs to build capacity of managing partners and local production and dissemination teams. Once local capacity is built, the need for international technical oversight is reduced significantly, which correlates with significant reductions in cost to continue implementing and scaling the approach regionally and nationally.

An estimation of the cost for large-scale expansion of the project will be conducted in coordination with partners, reflecting optional scenarios including: 1) a scale-up of the same community-video model in culturally similar Hausa-speaking communities in Maradi and Zinder using the 10 videos already produced; 2) a scale-up using the same model as option one, adding production and dissemination of additional videos on other resilience-related themes (agriculture, governance, WASH, emergency preparedness for response to shocks, etc.); and 3) an extension of the community video model to different regions of the Sahel, which would require formative research, adaptation of the training materials and monitoring tools, and cultural and language adaptation of video content.

Participatory Stakeholder Workshop

The objectives of the workshop session with program beneficiaries were to:

• share experiences and perspectives on the value of the community video approach, including whether and how the approach changed the dynamics of the groups meetings
• reflect on video acceptability, including differences between departments and groups stratified by participant gender and age
• solicit feedback on opportunities for scale-up and lessons
• discuss preference of video topics and ideas for future topics.

The objectives of the workshop session with program volunteers or mediators were to:

• share experiences and perspectives on the value of the community video approach
• reflect on video acceptability, including differences between groups stratified by participant gender and age
• solicit feedback on opportunities for scale-up and lessons
• discuss options for sustainability and local ownership of the approach
• discuss role of mediators, including impact on social status and challenges such as time use in the approach.

 The video gave us the final push for change. I already knew many things before when attending the écoles de maris, but it took us to see the video to make a change.
--Program beneficiary

Feedback from the workshops was promising. Community members, including WRA, men, and adolescents, felt that the videos were good communication tools in their context. Some beneficiaries indicated that these videos were the first they have ever seen. People related to and felt that they could recognize themselves in the videos, even in Aguié, where videos were only disseminated and not produced. The approach incited behavior change at the household level, but it also created social change as all stakeholders felt that it was important and useful to include men in the disseminations. There is a huge demand from other village members to be included in screenings, and also from other villages where the videos were not shown. Some communities even began to collect money to put continue the work after the program, if they were not allowed to keep the pico projectors and production equipment. Most groups want more screenings per month as well as more supportive home visits.

 We will establish a fund to support the costs related to the mediator. We want the mediators to continue doing their work.
--Program beneficiary

The following points should be considered in any phase-two program or scale up:

• Some type of payment for the mediators, since it has been a challenge for their efforts to be considered purely volunteer.
• Improvements in the conditions under which the interventions are conducted (e.g., providing mats to sit on).
• Meeting community demand for additional video screenings without compromising the important work of the established groups.

SPRING/DG team along with beneficiaries, mediators, and partners during the Participatory Stakeholder Workshop in July 2015 in Maradi, Niger

Conclusion

The SPRING/DG collaboration in Niger highlights the potential of community video as an innovative and effective way to foster social and individual nutrition and hygiene-related behavior change. This one year proof of concept in the Maradi region showed that this approach can be implemented successfully in the resilience context of Niger. Initial reports indicated that SPRING and DG were able to train local partner staff in community video production, facilitated dissemination, and data collection and management. The effectiveness of community video on two priority behaviors (handwashing and responsive feeding) has been evaluated and recommendations for scale-up made. The costs associated with adapting and implementing the community video approach, including projections about the cost of scaling to other communities and regions, have been defined. According to COCO, the facilitated video screenings resulted in the adoption and/or promotion of the high-impact nutrition behaviors that were featured. Preliminary findings indicate that the SPRING/DG collaborative approach empowers women, men, and communities to improve their health through improved nutrition and hygiene behaviors in a community-led process. The participatory stakeholder workshop provided evidence that the approach led to household level behavior change and social change. There is now increased demand for this approach in the study region and beyond.