Adapting the Social Norms Exploration Tool in the Democratic Republic of the Congo to Identify Social Norms for Behavior Change

Social Norms and Family Planning in the Democratic Republic of the Congo

Social norms are unwritten rules of behavior shared by members of a group or community. They dictate what people believe is normal (descriptive norms) and appropriate or approved (injunctive norms) behavior.¹ Power hierarchies that favor men over women are manifest in restrictive social and gender norms that negatively influence women’s health and well-being.² Research, globally and in the Democratic Republic of the Congo (DRC) specifically, indicates limited spousal communication about health and pervasive norms that position men as decision-makers, inhibiting women’s access to health services.^3,4 A secondary analysis of the 2013–2014 Demographic Health Survey in the DRC found that 54.1% of women surveyed said their partner alone was the decision-maker for the respondent’s health care,⁵ and that women who had no say in their health care were less likely to be users of modern contraception than women who made their own health care decisions.⁶

Men’s lack of communication around family planning contributes to social norms that discourage adoption of family planning. These social norms include perceptions that family and community members do not support family planning and that there is pressure to have more children. Systematic reviews of qualitative research on family planning and social norms in sub-Saharan Africa identify social support and gender norms as influencing factors. A review of 13 qualitative studies on contraceptive use among young people found that social support for contraception was a motivator for contraceptive uptake, while social norms against discussing contraception acted as societally based barriers.⁷ In a review of 23 qualitative studies of determinants of condom use among adolescents, 13 studies noted that gender norms affected the practice. Two studies in the review highlighted that negative peer norms discourage condom use among young men.⁸

The Social Norms Exploration Tool (SNET; https://www.irh.org/social-norms-exploration) was developed by the Institute for Reproductive Health at Georgetown University as a rapid assessment tool to aid program implementers in conducting qualitative, participatory activities to better understand the social norms that influence behavior in a community and that are of interest to their project. It includes 5 steps to explore social norms and inform interventions: (1) plan and prepare, (2) identify reference groups, (3) explore social norms, (4) analyze results, and (5) apply findings. Tools and templates, which are intended to be modified, are provided for each step.

The SNET was validated via pilot testing with 2 projects (Masculinité, Famille, et Foi and “Growing up GREAT!”) in Kinshasa, DRC, from 2015 to 2019, before being tested in other countries.⁹ From 2018 to 2020, the SNET was applied by the Passages Project in Burundi, Niger, and the DRC,^10–12 by Save the Children in Nigeria, Burkina Faso, and Palestine,^13–15 by the MOMENTUM project in Nigeria,^16,17 and by the U.S. Agency for International Development (USAID)-funded Tulonge Afya project in Tanzania.¹⁸ Of the SNET applications for which reports are available, most focused on gender norms around reproductive health, early marriage, or intimate partner violence. In these applications, the SNET was adapted for use in different ways. For example, to identify reference groups, most applications used the SNET’s “My Network Tool,” but other projects have used in-depth interviews, voting during focus group discussions (FGDs), and FGD notes. To explore social norms, several applications used vignettes as the participatory activity, often complemented by the “5 Whys” activity, whereas 2 applications implemented the problem tree analysis only.

This study adapted the SNET and sought to identify the social norms and reference groups influencing male engagement in family planning discussions, decision-making, and uptake of methods to inform programmatic interventions in the DRC. A secondary objective was to further validate the SNET by applying the tool in smaller cities and rural areas in the DRC. We describe the value added by the SNET and challenges to adapting the tool, as well as recommendations for future use as the basis for human subjects qualitative research.

Adaptation of the Social Norms Exploration Tool

The USAID-funded project Breakthrough ACTION DRC adapted the SNET for this qualitative research study. The study was implemented in 3 DRC provinces with low contraceptive prevalence representing varying regions, ethnic, and linguistic groups: Lualaba, Sankuru, and Kasaï Central.

Applying the SNET in the DRC involved the selection and adaptation of data collection exercises presented in the SNET and the development of data collection instruments.

Planning and Protocol Development

Following the SNET’s plan and prepare guidance for step 1, we conducted the problem tree analysis with project staff to analyze barriers to contraceptive use among married couples in the DRC and identify root causes of low adoption of modern contraceptive methods. This process, combined with the 5 Whys activity, narrowed down the behaviors and populations of interest and shaped the research questions. The plan and prepare guidance, along with steps 2 and 3 of the guide, informed the development of the study protocol and adaptation of data collection instruments (Table 1).

Data Collection

In each province, the sample included individuals from 1 health subdistrict within each of 2 urban and 2 rural health zones. Data were collected in French, Swahili, or Tshiluba by 3 teams of 4 data collectors from the local firm ALMA Research Services, and data collection took place in 3 phases between March 11 and April 1, 2022 (Table 2). Data collectors were assigned to province teams based on their comfort with the region and fluency in the local language. All data collectors had prior experience conducting qualitative research. All data collectors participated in a 5-day training with in-depth review of the study instruments, practice with all instruments and activities, including role-playing, and a module on social norms.

The first phase of data collection corresponded to the SNET step 2, identify reference groups. Adapted from the SNET “My Social Networks” activity, the survey was digitized to a tablet-based questionnaire to improve data quality and facilitate rapid analysis of the results (as opposed to tabulating by hand). Unlike the example in the SNET, participants were asked only 1 question and were only able to share 1 influential person with whom they felt comfortable discussing family planning methods. In addition, rather than the SNET approach of selecting fewer participants through convenience sampling, we selected a larger sample through systematic household-based sampling (Table 1). Men (n=317) identified the people in the community with whom they were most comfortable discussing family planning (Table 2).

The second phase of data collection mapped to the SNET step 3, explore social norms. Although the participatory activities in the SNET are designed to be facilitated with groups, the vignettes can be implemented individually, as we did in this adaptation. Semistructured, individual interviews were conducted with a subsample of married and unmarried men (n=24) who had participated in the rapid questionnaires (Table 2). Interview participants were asked open-ended questions about their experiences with and involvement in family planning and their perception of men’s role in this process. In the interviews, a vignette describing a couple contemplating using family planning was used to examine how men in their community would respond to specific situations around birth spacing, couple communication, and permitting their partner to adopt a modern family planning method.

The third phase, a continuation of exploring social norms (step 3), took place several days later. Data collectors analyzed the questionnaire results from the first phase and recruited participants matching the reference group criteria to focus groups. The focus group discussions (FGDs) were designed to understand how reference groups reinforce descriptive and injunctive norms around family planning. A total of 144 participants participated in 24 FGDs, which included the SNET’s 5 Whys activity and a shortened version of the vignette (Table 2). In the DRC adaptation, we chose to facilitate the 5 Whys activity in small groups rather than in pairs, with 1 facilitator assigned to each group. This decision reflected the complexity of the activity in simulations during the data collector training.

Data Analysis

We adapted the analysis phase from SNET step 4 to accommodate a more conventional qualitative data analysis. While the SNET provided note-taking templates, we supplemented note-taking with audio recordings of qualitative data collection activities, subsequently transcribed into French, to preserve data not captured in the notes. Given the length and number of transcripts, we inductively coded a subset of the transcripts with qualitative analysis software ATLAS.ti and grouped the codes into 6 overarching themes.

Analyses of FGD and interview transcripts occurred through a participatory 5-day workshop that convened data collectors, program staff, and Ministry of Health representatives. The 6 themes were shared with the workshop participants for validation and the addition of any other reoccurring themes based on their reading of the transcripts. Specifically, the participatory analysis workshop included the following steps. Each participant was asked to read 2 transcripts and to discuss their findings with another participant who had been assigned transcripts from the same province and geography (urban/rural). Then, teams of 4 were assigned to read transcripts from each province. The teams came together to discuss similarities and differences between the urban and rural areas in their assigned province and engaged in the development of subthemes (with supporting citations). Finally, each team presented their findings and recommendations for their assigned province in plenary, and the presentations, notes, and supporting citations were shared with the research team.

Ethical Approval

The study was approved by the Institutional Review Boards of the Johns Hopkins Bloomberg School of Public Health (IRB approval #18667) and the University of Kinshasa School of Public Health (IRB approval # ESP/CE/04/2022). Data collectors were trained on human subjects research. All participants gave their informed consent.

Assessment of Adapting the Social Norms Exploration Tool for Research

We found it useful to develop a set of criteria with which to evaluate our experience adapting the SNET methods for social norms research. The criteria for analyzing the utility of the SNET-recommended methods make reference to the stated aims of the 5 steps:

1. How did the SNET guidance and activities facilitate development of the research questions, protocol, and instruments?

2. How completely were reference groups identified during phase 1 of data collection?

3. How successful was the adoption and adaptation of recommended participatory activities for exploring social norms during interviews and FGDs?

4. How did SNET results inform development of the data analysis workshop?

5. How effectively did results from the social norms exploration translate into recommendations?

Phase 1: Identification of Reference Groups

In phase 1, a rapid questionnaire was applied to identify reference groups. Given that the results of phase 1 were the basis for recruiting focus group participants in phase 3, we wanted to avoid the bias that could result from talking to a small convenience sample of men and extrapolating their nominated reference groups to the entire community. Thus, 317 men across the 3 provinces were recruited door-to-door for participation in the questionnaire. Seventy-five percent of the men were married, and all men were aged 15–39 years (Table 3). Results from these questionnaires informed the selection of reference groups in each province.

We asked men to identify with whom they felt most comfortable discussing family planning. For each province and geographic site (urban/rural), we selected reference groups based on the 2 most-cited responses. Table 4 shows the most-cited group, health care providers (47.0%), followed by friends (24.0%) as the second most-cited reference group, and family members (20.2%).

Phases 2 and 3: Exploration of Social Norms

Phases 2 and 3 of the study explored social norms through participatory activities. In Sankuru, religious leaders were not among the most cited groups but were selected for participation due to the emerging theme in the interviews around the importance of religion. Themes from the qualitative data collection activities represented normative factors as well as individual factors, such as knowledge and attitudes (Table 5).

The results highlighted both descriptive and injunctive norms related to birth spacing, uptake of modern methods, and couples’ communication and decision-making. Birth spacing was a common and approved practice, albeit frequently practiced through traditional methods rather than modern methods. Gender norms invoking men as heads of household gave men authority over their wives and the final say in decisions about birth spacing and family planning method use. Injunctive norms stipulated that men who ceded decision-making power to their wives or whose wives used family planning were seen as weak or impotent. Health care providers who prescribed contraception without the husband’s consent could also face consequences, including the husband bringing a complaint against them.

At the individual level, respondents recognized the health benefits of birth spacing for mothers and children, as well as the economic benefits. Many men expressed concerns about allowing their wives to adopt a modern method or to select the method of her choice. They cited fears around women’s infidelity, negative side effects, and infertility.

Finally, findings reflected the importance of reference groups on male engagement in family planning and the distinction between reference groups and sources of knowledge. Overall, men were most comfortable talking to health care providers about family planning. Close friends served as both sources of knowledge and reference points for descriptive norms. Religious leaders were not cited as often, but religious mores that frame family planning as sinful had clear influence on injunctive norms.

Phase 4: Analysis

This study expanded beyond the SNET guidance provided for the analysis step to ensure that all the data were captured, rather than relying on note-takers, which could have resulted in inconsistencies across study activities. Because the data collectors did not have a background in social norms before their training for this study, we wanted to ensure that the note-takers did not miss or gloss over potential social norms for lack of recognizing them as norms. During the participatory workshop, participants read the transcripts and discussed emerging themes and insights, first by site, then by province, and finally in plenary.

Phase 5: Recommendations

Results and recommendations from the participatory workshop were shared through dissemination events in the study provinces. We proposed recommendations to apply the study findings to the development of interventions. We hosted workshops with local health officials to translate findings into program activities that could be added to operational action plans. Deviating from SNET Step 5 (application of findings), we followed a multistep framework to bridge the gap from study findings to evidence-based program planning. Results in the form of themes and exemplary quotations, as well as recommendations by reference group, were shared via dissemination events that were followed by a data use workshop intended to adapt programmatic messaging and generate province-level health communication activities. We incorporated questions from SNET Step 5 into the workshops to help health officials address social norms with their workplan activities.

Step 1: Plan and Prepare

The SNET provides options for exploring social norms, with limited guidance on how to select the methods. This application of the SNET was among the first to be conducted as a formal research study undergoing ethical review. The SNET does not provide guidance on processes for ethical review of research protocols or sampling appropriate for a qualitative research study conducted across different sites and provinces. While the SNET states that implementers should “hold a team discussion on ethical considerations of the social norms exploration,” it does not discuss obtaining approval from ethical review boards nor anonymity of data. The Breakthrough ACTION DRC project approached the research systematically, in part, due to the need to submit a full protocol to the Institutional Review Board and local ethics committee according to USAID policy.¹⁹ Utarini et al. list 11 “critical criteria” for evaluating rapid assessment procedures, of which 1 is around ethics and the need to obtain approval from ethical review boards. The lack of explicit recommendations for ethical review and approval in the SNET is a weakness, albeit one that is common across similar rapid assessments.²⁰

Based on our experience, we recommend the SNET for conceptualizing social norms research, with specific recommendations for adaptation. For example, the team conducted the problem tree analysis (Table 6) during the planning period. This brainstorming methodology gave surface-level responses rather than root causes behind non-use of family planning methods. Instead, we recommend asking project staff to communicate their experiences or observations of the behavior of interest through narrative storytelling and then elicit root causes from those examples. Similarly, the 5 Whys activity (Table 6) did not result in a deeper understanding of root causes.

Step 2: Identify Reference Groups

The reference group questionnaire allowed for rapid analysis of reference groups and precise tailoring of reference group criteria by site. One challenge is the assumption that those with whom men are most comfortable discussing family planning are those from whom men receive cues about social norms. In the study, men also received cues about social norms from other sources with whom they were less comfortable, such as religious leaders who may reinforce injunctive norms about family planning as sinful. SNET implementers should attempt to take other normative influencers into account.

The SNET and associated guidance stress the importance of identifying specific attributes of reference group members. If a man were to nominate a cousin, follow-up questions would identify the cousin as male or female, maternal or paternal, older or younger. This level of detail was relevant to SNET applications in which data collectors would recruit the specific person cited, as in Nigeria.¹³ However, it did not serve to define categories of reference groups (i.e., older, maternal, male cousins). From a practical standpoint, such a specific reference group is not functionally different from “male peers aged 18–49 years.” Rather than building the questionnaire around 1 question and asking follow-up questions about the specific person cited, future applications should include questions to ascertain who is influential in supporting the behavior of interest, as well as which community members are not supportive. Dichotomizing the reference groups into “enforcers” of sanctions versus “social supporters,” as in Burundi,¹⁰ would help researchers better understand the social pressures on the population of interest.

Step 3: Explore Social Norms

FGDs with reference group members were important for understanding the social context, exploring different perspectives, and triangulating findings between groups. Qualitative data from married women, religious leaders, and health providers created a fuller picture of the pressures and norms surrounding family planning access. FGDs identified how reference groups influence men’s involvement in family planning and decision-making.

Vignettes (Table 6) were incorporated into the individual interview and FGD guides. The vignettes represented scenarios that participants recognized from their communities and provided a way for respondents to discuss sensitive issues without speaking to personal experience. However, providing too many plot points was confusing and made it difficult for respondents to understand how each series of questions differed from the series before. The SNET recommends that reference group members be introduced in the vignette. In Nigeria, vignettes included fill-in-the-blanks to allow participants to direct the story in the way they felt was most plausible (“She goes to speak to _____”). This example provides a good option for future applications to capture the role reference groups play.

The second SNET-recommended activity applied only in FGDs was the 5 Whys activity (Table 6), characterized as an activity of easy-to-medium difficulty. While easy to develop, it was confusing to implement. Rather than asking a question and delving further into root causes with each additional “why,” participants gave different, surface-level responses to the initial question. Participants were confused about how each “why” was distinct from the previous iteration. The time spent on the 5 Whys exercise did not add value to the findings, and we do not recommend it to others interested in using the SNET.

Step 4: Analyze Results

In the “Analyze Results” section of the SNET, data are compiled based on activity templates and notes taken during data collection. Our challenge was moving from raw transcripts to a stage where the data had been sufficiently distilled such that it was possible for the team to apply the SNET guidance for analysis. Given the number of transcripts and the desire for shared ownership of the results, we arranged for collaboration between different stakeholders and intercoder discussion to compare findings across sites. A subset of qualitative transcripts was inductively coded before the workshop, and validation of existing themes and addition of new themes occurred at the end of each day of the workshop. A collaborative synthesis was conducted by small groups corresponding to each province and shared with the rest of the participants at the end of the workshop.

The concept of social norms was not intuitive for the workshop participants. Many participants did not have a deep understanding of social norms, and a brief training was insufficient to prepare the participants to fill out the SNET example templates to rank the relative influence of normative factors. Certain concepts were misunderstood, such as social sanctions as defined in the social norms literature versus negative consequences more generally (for example, experiencing side effects after adopting a family planning method). Despite the SNET focusing on social norms, the participants did not find that social norms were the most important influence on behavior. Many recommendations from the workshop focused on increasing knowledge rather than addressing normative factors. Future applications of the SNET should include activities to better acquaint workshop participants with social norms and provide sufficient time to synthesize norms from the data. One of the key findings reflected in other articles on rapid assessment procedures is the need for capacity-building in social sciences to go hand-in-hand with provision of ethnographic or qualitative research manuals,²¹ with 1 article noting that “general capacity building in the social sciences is necessary to assure the validity and appropriate use of program-specific manuals.”²² Spending more time during the data collector training to ensure that all data collectors had a good understanding of social norms and how they operate and doing a short refresher before the data analysis workshop would have better equipped data collectors to focus on identifying social norms rather than other behavioral influences. The unspoken nature of social norms and greater familiarity with other behavioral drivers, such as knowledge and attitudes, presented a challenge to identifying different norms and their function in the community.

Step 5: Apply Results and Formulate Recommendations

The concluding section of the SNET presents questions for the program team to (1) adjust project design and (2) adjust research, monitoring, and evaluation. The questions provided in the SNET for adjusting project design helped inform the microplanning work by asking participants to identify obstacles based on the findings and identify how to include influential groups in future interventions. Questions, such as “What positive norms can we build on?” “What harmful norms should we address?” and “Which people and (reference) group are influential?,” helped to structure the conversation about adapting the findings from the social norms exploration into activities that could be added to future workplans. Members of the provincial health departments expressed that they appreciated that the research team returned to their provinces to share the results and think about how to apply them, although the SNET guidance says to do this process as a program team rather than with local health officials.