The effect of radio dramas on willingness to report intimate partner violence

Introduction: Intimate partner violence remains widespread partly because survivors and witnesses are reluctant to report incidents to authorities.

Methods: We examine whether narrative media messages can increase willingness to report intimate partner violence (IPV) in rural Tanzania through a survey experiment involving 1,009 respondents across 15 villages. Participants were randomly assigned to hear a 1.5 min radio drama or to a pure control condition.

Results: Fewer than 10% of control group respondents recommend that IPV survivors report to police, the authority respondents believe is most likely to punish the abusive husband, but also most likely to cause the husband to abandon the survivor. Treated respondents were 8.9 percentage points more likely to recommend that survivors report to the police (p < 0.001). However, the drama did not improve other pre-specified outcomes—willingness to report to village chair-persons or witnesses reporting IPV—nor did it affect underlying attitudes about IPV acceptability, perceptions of community norms, or beliefs about reporting consequences.

Discussion: Our findings highlight the distinction between citizens’ views of police and local political authorities as targets of IPV reporting. The results also suggest that narrative dramas can influence behavioral intentions central to their narrative, but with limited spillover to related attitudes or behaviors.

1 Introduction

Intimate partner violence (IPV) threatens women’s health and well-being worldwide. One in four women aged 15–45 will experience physical or sexual abuse from an intimate partner in their lifetime [World Health Organization, 2023]. In Tanzania, where our study takes place, 44% of women and girls aged 15–49 experience IPV in their lifetime and 24% experience it in a given year (World Bank, 2022).

IPV persists in part because both survivors and witnesses are reluctant to report incidents to authorities (McCleary-Sills et al., 2016; Katiti et al., 2016). Failure to report stems from acceptance of IPV, distrust of authorities, and fears about familial or social backlash (Simmons et al., 2020). But an important dimension of IPV reporting remains poorly understood: to whom do survivors and witnesses choose to report, and why? In settings with weak state capacity and legal pluralism, citizens face consequential choices between reporting IPV to formal law enforcement and local political authorities. These institutions inspire different levels of trust, impose different barriers to access, and produce different outcomes for survivors. Existing research has treated “reporting to authorities” as unitary behavior, eliding these distinctions (Blaydes et al., 2025).

One promising tool for promoting IPV reporting is narrative media. A growing body of evidence suggests that narrative entertainment, or “edutainment,” can shape preferences around social issues such as public health (Green et al., 2021; Krause and Fletcher, 2025), environmental protection (Rahmani et al., 2024), violence prevention (Bilali, 2022), and gender empowerment (Montano et al., 2023; Green et al., 2023). A recent meta-analysis finds that narrative messaging influences attitudes, behavioral intentions, and behaviors by attracting audiences and modeling preferred behaviors (Rahmani et al., 2025).

In the context of IPV, narrative dramas have been shown to shape both cognitive attitudes about the acceptability of IPV (Banerjee et al., 2019; Fotini et al., 2023), as well as conative attitudes towards witness reporting of IPV (Green et al., 2020; Cooper et al., 2020). However, cumulative evidence about the edutainment interventions’ effects on IPV remains mixed (Peterman, 2025; Arias, 2019; Hoff et al., 2021; Filippo et al., 2023; Yount et al., 2023; Clark et al., 2024). Evidence is especially sparse with respect to brief, one-act narrative dramas that are common parts of many public health campaigns. Whether short-duration dramas reliably shift attitudes, perceived norms, and behavioral intentions remains unclear.

The purpose of this study is to twofold. First, we use embedded vignette experiments to understand rural Tanzanian’s beliefs and intentions about whether to report IPV, but also to whom they should report. Second, we use a lab-in-the-field experiment to evaluate whether a short, targeted radio drama can influence citizens’ beliefs about whether to report IPV as well as to whom to report IPV.

We randomly assigned 1,009 respondents across 15 coastal villages in Tanzania’s Tanga region to hear either a 1.5-min radio drama or no audio in the context of a 45-min survey. The drama, designed by a local NGO as part of a broader initiative to improve government responsiveness to IPV, aimed to address three barriers to reporting: fears that police would not believe survivors, concerns about family punishment, and worries about community support. In the drama, a young woman reports her abuse to police after friends describe their positive experiences with police responsiveness and community support.

After the intervention, enumerators used vignette experiments to elicit respondents’ views about the acceptability of IPV, appropriate responses to IPV by survivors and witnesses, and the likely reactions to a survivor reporting IPV by the survivors’ husband, in-laws, neighbors, and local authorities (Bicchieri, 2016; Cooper et al., 2020). Our lab-in-the-field design positions us to answer three questions: (1) What are existing attitudes, perceived norms, and beliefs around survivor and witness reporting of IPV to different authorities? (2) Can a short narrative radio drama influence support for IPV survivor and witness reporting? (3) Can a short drama influence three possible drivers of survivor and witness reporting: views about the acceptability of IPV, perceptions of community norms towards IPV behaviors, and beliefs about the consequences of IPV survivor reporting?

To preview results, we find widespread resistance to survivors and witnesses reporting IPV, especially to the police. We find limited evidence for our primary pre-specified hypothesis that the drama would increase survivor and witness willingness to report to any state authority. Instead, we find strong evidence for a treatment effect only on the behavior that was explicitly modeled in the drama: survivors reporting to the police. We also find no evidence that the radio drama influenced hypothesized mechanisms: respondents’ underlying acceptance of IPV, perceptions about the frequency of survivor reporting, or beliefs about the responses to IPV reporting by the survivors’ family, community, and government authorities.

Our findings contribute to the literature on both IPV reporting and entertainment-education. First, we highlight an important distinction that has not previously received sustained scholarly attention. When citizens report abuse in communities with weak state capacity, to whom they choose to report may be as consequential as whether they report at all (Simmons et al., 2020; Lazarev, 2023). We show that Tanzanian villagers hold sharply different beliefs about the consequences of reporting to formal law enforcement versus local political authorities, and that an intervention that increases support for reporting to the former has zero-to-negative effects on prioritization of the latter.

Second, the study advances research on narrative persuasion by evaluating whether brief narrative dramas can influence conative attitudes toward IPV reporting (Montano et al., 2023; Green et al., 2020). We show that even a 1.5-min radio spot can shift beliefs about appropriate responses to IPV. At the same time, our results underscore the limits of short narrative interventions. The effects of the drama are narrow: changes in conative attitudes are confined to the specific behavior modeled in the narrative—survivors reporting abuse to the police—and do not reliably extend to closely related behaviors, such as reporting to village chairpersons or witness reporting.

Third, our null findings on mechanisms illuminate both how brief narratives work and how practitioners might use them more effectively. We find no evidence that the drama alters attitudes about IPV acceptability, perceptions of community norms around reporting, or beliefs about reporting consequences. This suggests that short narratives operate primarily through direct behavioral modeling rather than by reshaping the attitudes and beliefs that typically motivate such behaviors. For practitioners, it suggests that effective campaigns require multiple message formats: diverse dramas to model a range of behaviors alongside complementary content to address underlying attitudes and norms.

1.1 Literature review

1.1.1 Intimate partner violence

Intimate partner violence represents one of the most under addressed threats to women’s health worldwide (Blaydes et al., 2025). In lower and middle-income countries (LMICs), roughly one in three women will experience physical or sexual violence from an intimate partner during her lifetime (World Bank, 2022). The consequences of IPV extend well beyond physical injury. Survivors face elevated risks of depression, anxiety, and post-traumatic stress disorder (Lappeman et al., 2025; Dillon et al., 2013; Yonga et al., 2022); they are more likely to miss work or leave the workforce entirely (Vyas et al., 2023); and they often experience diminished autonomy within their households (Yonga et al., 2022). Sub-Saharan Africa bears a disproportionate share of this burden, with cross-national analyses consistently ranking the region among the highest in IPV prevalence globally. Approximately 36% of ever-partnered women have experienced physical or sexual violence from an intimate partner (Mossie et al., 2023).

Tanzania exemplifies the severity of this crisis. National survey data indicate that approximately 44% of Tanzanian women and girls have experienced physical or sexual IPV, with prevalence varying by region, education, and household wealth (World Bank, 2022). Rates are particularly high in rural and low-income communities, including the villages where this study takes place (Nabaggala et al., 2021).

A growing interdisciplinary literature explores the drivers of IPV (Blaydes et al., 2025; Peterman, 2025; Ellsberg et al., 2015; Bourey et al., 2015). In Heise (1998)’s seminal ecological framework, IPV emerges from a complex interaction of personal histories and biological factors, power imbalances within families, peer influences shaped by societal and religious norms, and broader patriarchal gender ideologies (Mootz et al., 2023). Research in Tanzania has further emphasized alcoholism, attitudinal acceptance of IPV, and women’s lack of independent income as contributing factors (Abramsky et al., 2019; Reese et al., 2021).

These multifaceted causes of IPV suggest multiple potential intervention strategies for reducing its incidence. Common approaches in LMICs include economic empowerment programs (del Eggers Campo and Steinert, 2022; Haushofer et al., 2019), reforms to laws and law enforcement practices (Sanin, 2021; Ferraz and Schiavon, 2022), and campaigns to shift attitudes and perceived social norms (Kapiga et al., 2019; Green et al., 2023). Some scholars and practitioners advocate for holistic approaches that combine economic, legal, community-based, and media interventions (Lees et al., 2021).

Our study addresses two gaps in this literature, which we discuss in more detail below. First, while many interventions have focused on reforming laws and improving formal responses to IPV, fewer studies have addressed the underlying social and attitudinal barriers to IPV reporting (Umubyeyi et al., 2016). Second, while holistic programs have proven effective in several contexts, our study helps isolate the effects of a program component: short-form media messaging (Green et al., 2020).

1.1.2 Intimate partner violence disclosure and reporting

An understudied driver of IPV’s persistence is the presence of widespread norms against disclosure and help-seeking by both survivors and witnesses (Felson and Paré, 2005; Felson et al., 1999; McCleary-Sills et al., 2016). The cross-national analysis estimates that only 7% of women who experience IPV report to any formal source, suggesting that the true prevalence may be 14 to 44 times higher than the capture of official statistics (Palermo et al., 2014).

Despite growing IPV response and prevention programming in Tanzania, survivors and witnesses rarely report incidents to authorities (Montano et al., 2023; Katiti et al., 2016). This reluctance stems from multiple sources: beliefs about the underlying acceptability of IPV (Rohn and Tenkorang, 2023); distrust of police and formal legal processes (Jassal, 2024; Jassal, 2020; Aiko, 2015; McCleary-Sills et al., 2016); and fears about social and familial repercussions (Cooper et al., 2020). These norms help explain why legal reforms alone have had limited impact on IPV incidence, particularly in low-income areas (Umubyeyi et al., 2016). While Tanzanian women’s rates of help-seeking exceed those in some other African countries, significant barriers remain in rural and low-income regions (Dickson et al., 2023).

When survivors do report, they face consequential choices about which authority to approach. In settings with legal pluralism and varying state capacity, citizens may turn to formal law enforcement, local political representatives, traditional leaders, or family networks. In rural Tanzania, village chairpersons (“mwenyekiti”) serve as the first point of contact for many disputes, offering a more accessible alternative to distant police stations (Katiti et al., 2016; Simmons et al., 2020). Research from Nigeria indicates that only about 1% of women willing to disclose IPV would report to police, with most preferring family members or religious leaders (Okenwa et al., 2009). These patterns suggest that research about IPV reporting must do more to grapple with citizens’ differentiated perceptions of local versus formal authorities.

In addition to addressing the distinction between disclosure to formal versus informal authorities, our study builds on existing research by moving from documenting the causes of non-disclosure toward testing interventions designed to increase survivors’ and witnesses’ willingness to report.

1.1.3 Entertainment education

Narrative entertainment, or “edutainment,” has become an increasingly common approach to influencing social attitudes and norms (Paluck and Green, 2009; Banerjee et al., 2019; Murrar and Brauer, 2018; Blair et al., 2019). Wang and Singhal define edutainment as a “theory-based communication strategy for purposefully embedding educational and social issues in the creation, production, processing, and dissemination process of an entertainment program, in order to achieve desired individual, community, institutional, and societal changes among the intended media user populations” (Wang and Singhal, 2021).

Several theories explain why narrative dramas may be particularly influential. Social Cognitive Theory (SCT) posits that dramas promote change by providing audiences with compelling models of targeted behaviors (Bandura, 2004). The Extended-Elaboration Likelihood Model (E-ELM) suggests that narratives shift attitudes by absorbing audiences in stories, thereby reducing their tendency to counter-argue against unfamiliar messages (Slater and Rouner, 2002). Other theories of narrative persuasion highlight that information embedded in narratives is more influential because it is easier for audiences to understand (Bullock et al., 2021). More simply, narratives may attract and retain audiences who would otherwise prefer purely entertainment media, providing an effective vehicle for prosocial content (Rahmani et al., 2025).

Experimental evidence supports these theoretical propositions. A recent meta-analysis finds that narrative interventions generate consistent effects on knowledge, attitudes, and behavioral intentions (Rahmani et al., 2025), with similar patterns observed in meta-analyses focused specifically on health outcomes (Doré et al., 2025).

Research also suggests that edutainment may be a powerful tool for reducing IPV (Banerjee et al., 2019; Green et al., 2020; Hoff et al., 2021; Fotini et al., 2023). In a systematic review of 21 edutainment programs, Peterman (2025) finds that thirteen interventions in developing countries positively influenced attitudes toward violence against women. Studies closely related to ours have found encouraging effects of narrative dramas on IPV reporting intentions in East Africa (Green et al., 2020; Montano et al., 2023).

1.1.4 Gaps in existing literature

Despite this evidence base, important questions remain. First, most studies evaluate multi-episode series through repeated screenings. However, media campaigns in practice are usually dominated by short, targeted dramas lasting 2–5 min. Whether brief, targeted messages can shift attitudes toward IPV reporting remains unclear. Short radio spots represent a more scalable approach than long-form programming, but their efficacy has not been rigorously tested.

Second, prior research has treated “reporting to authorities” as a unitary outcome without distinguishing between recipients. Yet citizens hold distinct views of police versus local leaders and face distinct barriers when considering each option. Our study draws attention to citizens’ differentiated views on reporting to different state authorities.

Third, the mechanisms through which edutainment operates remain incompletely understood. Theoretical accounts emphasize effects on attitudes about IPV acceptability, perceived norms around reporting, and beliefs about disclosure consequences. Existing studies have rarely measured all three within a single design.

1.2 Study context

The study takes place in the coastal district of Mkinga in Tanzania’s northeastern Tanga region.

Tanga provides an apt context for the study of intimate partner violence (IPV) for four reasons.

First, rates of IPV in Tanga are a significant concern. 31% of women and girls aged 15–49 have experienced physical, sexual, or emotional violence committed by their husband/intimate partner at some point in their lifetime (Tanzania Ministry of Health, 2023), higher than the global average (27%) (World Bank, 2022). Our survey results suggest that the prevalence of IPV is sustained in part by a culture of acceptance and non-reporting. Just 40% say that a survivor should report abuse to a formal authority, and only 33% say that they would report IPV to an authority if they witnessed it occurring in their community. Tanzania has recently initiated a 5-year National Plan of Action (NPA) against gender-based violence (United Republic of Tanzania, 2024). The NPA specifically targets the media sector for further investment, recognizing its broad reach and potential to shape social norms and values.

Second, rural Tanzanians face a variety of social and state-based options for responding to IPV. There are strong social pressures to resolve intra-household conflict within the family or through religious or traditional means rather than appealing to state authorities (Katiti et al., 2016; Simmons et al., 2020; Aiko, 2015; Simmons et al., 2020). When citizens do choose to turn to the state, they can appeal to village chairpersons (“mwenyekiti”), elected but unpaid village leaders who are responsible for informal dispute resolution, or they can appeal to formal law enforcement. Decisions about responding to intimate partner violence are deeply gendered and embedded in local hierarchies of authority (Katiti et al., 2016; Christopher et al., 2022). Women face substantial social and economic risks when reporting abuse, including abandonment by husbands, retaliation from in-laws, and community stigma, which shape perceptions of which responses are feasible or appropriate.

Third, the rapid growth of the radio sector in rural Tanzania makes it an ideal setting for testing the influence of radio dramas on social attitudes and behaviors (Green et al., 2024; Rahmani, 2025a). Research has shown that norms can be driven by leading those with strong opinions, such as community members, village leaders, and religious leaders (Tankard and Paluck, 2016; Nasir and Naaz, 2024). In East Africa, religious and community leaders have proven powerful voices for reducing intimate partner violence (Boyer et al., 2022). These distinct cultural and religious norms, combined with Tanzania’s reliance on radio for news and messages, show how rural Tanzania presents an important opportunity for an understanding of how media can influence societies with social hierarchies.

Finally, radio messages in rural Tanzania operate within a communication culture embedded in gender and community power relations (Sturmer, 1998). Radio is a familiar and trusted medium for conveying public messages about health, politics, and family life (Groves, 2023). Radio is commonly used by both state actors and non-governmental organizations to model socially appropriate behavior (Rahmani, 2025b; Green et al., 2021). However, local radio has historically been dominated by male voices, and women have been less represented by talk-radio programming (Montano, 2025). In recent years, several Tanzanian non-governmental organizations have successfully used radio narratives to disseminate more gender-progressive messages. These stations have earned substantial audience share and demonstrated an ability to engage female listeners (Green et al., 2024).

2 Materials and methods

2.1 Conceptual model and hypotheses

Our primary outcome of interest is respondents’ willingness to report intimate partner violence (IPV) to authorities. We focus on three reporting behaviors: victims reporting abuse to the police, victims reporting abuse to village chairpersons, and witnesses reporting IPV to authorities.

We hypothesize that exposure to the drama may influence respondents’ willingness to report IPV via indirect and direct pathways. Indirectly, the drama may shift attitudes about the acceptability of IPV, perceptions of community norms surrounding IPV reporting, or beliefs about the likely conse-quences of reporting abuse. Directly, the drama may influence behavior by directly modeling certain action (Bandura, 2001) independent of broader attitudinal change.

2.1.1 Reporting intentions

We begin with hypotheses about reporting behavior itself. The radio drama explicitly depicts a survivor choosing to report her abuse to the police. We therefore hypothesize effects on both the modeled behavior and two closely related reporting behaviors.

Hypothesis 1a: Treated respondents will express greater willingness to report IPV to the police if they experience abuse.

Hypothesis 1b: Treated respondents will express greater willingness to report IPV to the village chair-person if they experience abuse.

Hypothesis 1c: Treated respondents will express greater willingness to report IPV to authorities if they witness abuse.

2.1.2 Attitudes toward IPV

One pathway through which narrative exposure might influence reporting intentions is by changing attitudes toward the acceptability of IPV itself. If the drama reduces acceptance of IPV, respondents may become more willing to take action against it.

Hypothesis 2: Treated respondents will be more likely to say that IPV is unacceptable.

2.1.3 Perceived community norms

The drama may also influence perceptions of descriptive norms surrounding IPV reporting. In particular, it may shift beliefs about whether other community members would report abuse if they were victims and whether witnesses would support survivors by testifying on their behalf. When individuals believe that reporting is socially supported, they may be less fearful of social sanction.

Hypothesis 3a: Treated respondents will be more likely to believe that IPV victims in their community would report abuse.

Hypothesis 3b: Treated respondents will be more likely to say that they would testify on behalf of an IPV survivor who reports abuse.

2.1.4 Beliefs about reporting consequences

Alternatively, the drama may change beliefs about the expected consequences of reporting IPV. Based on prior qualitative interviews, we focus on four perceived outcomes: whether authorities will believe survivors, whether authorities will act on reports, whether survivors will face retaliation from husbands or in-laws, and whether the community will offer support.

Hypothesis 4: Treated respondents will be more likely to believe that reporting IPV leads to positive consequences for survivors.

2.1.5 Behavioral modeling

Finally, the drama may operate through direct behavioral modeling. Rather than altering beliefs or attitudes, listeners may simply adopt the specific behavior depicted in the narrative—reporting IPV to the police—consistent with social cognitive theories of learning from observed behavior (Bandura, 2001).

We measured outcomes corresponding to each hypothesized pathway. This design allows us to assess not only whether the radio drama increases willingness to report IPV, but also whether any observed changes are mediated by shifts in attitudes, perceived norms, or beliefs about reporting consequences, or instead reflect a more direct modeling effect.

2.2 Intervention

The intervention is a 1.5-min-long audio radio spot set in Tanga region, developed and produced by a non-governmental organization focused on gender justice. The radio spot was part of a larger media and community outreach campaign and government initiative designed to mobilize community and state responses to gender-based violence, including the recruitment of female police officers. It was designed to raise awareness about these efforts to improve IPV responsiveness by district police. The radio spot had been aired by a local radio station in the nearby Pangani region, but not in the villages where the study took place.

The radio spot was tailored to speak to local concerns about reporting IPV: that police will not believe the survivor; that the survivor’s husband and his family will punish the survivor; that community members would not support the survivor’s claims; and that the survivor’s in-laws would punish her. The radio spot is centered around a conversation between a woman and her friends. The audio clip follows the story of a woman who is being abused by her husband, but is afraid to act. The woman’s friends remind her that women in similar places reported their situation to the police, and the police believed the survivors and took action on their behalf. The protagonist of the story also learns that her friends would testify in court on her behalf if she decided to pursue action against her husband and the situation escalated to the court. After the conversation with the woman, the protagonist decides she is going to report her abuse to the police. A full script of the radio spot can be seen in Appendix A2.

The listening setting was in or near the respondent’s home, where the survey took place. Treatment group respondents listened to the drama on an individual basis as it played directly from the tablet that the enumerators used to collect data. There were no reports of audio-player problems. Subjects in the control group did not listen to any audio clip; the survey in the control group progressed directly to the outcome questions.

2.3 Sampling of villages

To ensure focus on rural communities, villages were selected if they were coastal, not located along a main road, and contained fewer than 500 households in the primary village. Figure 1 presents a map of participating villages.

Figure 1

Figure 1. Village sample.

2.4 Sampling of respondents

Households were randomly selected for participation in the survey from a village list. Table 1 shows the individual-level characteristics of our sample in the treatment and control conditions. For a full list of baseline characteristics, see Table A1.

Table 1

Table 1. Sample description, by treatment and control group.

The average age of respondents was 39 years old. 70% of the sample was married, and the average number of children per household was 2. The study sample is almost exclusively Muslim, consistent with the dominance of Islam along Sub-Saharan Africa’s Swahili coast. While this represents a potential scope condition for the study, similar studies have found consistent effects of media messages about IPV for Muslim and Christian respondents (Montano et al., 2023).

Turning to economic conditions, the four most common self-reported forms of employment are fishing (27%), farming (19%), small-business ownership and self-employment (33%), and household (13%). On the whole, villagers in coastal Mkinga are not well off: just 40% have access to the electricity grid and less than half live in homes with concrete floors.

Respondents were also interested in news and familiar with radio. Almost all respondents indicated that they are somewhat (41%) or very (43%) interested in news, more than 50% have listened to the radio in the past week, and more than 90% have listened to the radio in the past year.

2.5 Random assignment

We conducted simple random assignment to the experimental condition at the individual level using SurveyCTO. After subjects completed baseline questions, the software automatically assigned subjects to the treatment or control group with equal probability and without stratification. 498 were assigned to the treatment group and 511 were assigned to the control group. Table 1 and Table A1 show the distribution of baseline covariates in the treatment and control conditions; as expected, random assignment was not significantly correlated with any pre-treatment characteristics of respondents.

2.6 Data collection

Data was collected by gender-matched enumerators in a private, one-on-one setting between February 1 and February 21, 2025. All participants provided both oral and written consent. The treat-ment was embedded in a larger survey that averaged 45 min to complete. After asking a series of baseline questions, enumerators played the audio clip to treated respondents and then proceeded immediately to the outcome questions described below.

2.6.1 Outcome measures

We collected four categories of outcome measures: attitudes towards reporting IPV, attitudes about the acceptability of IPV, perceptions of community responses to IPV, and beliefs about the consequences of IPV reporting.

To measure attitudes and perceived norms, enumerators first read a short vignette to respondents: “Now I am going to tell you a short story about a woman in a nearby village, and then ask you some questions about what you think the woman should do. Imagine that a [rich/poor] woman in your village was [beaten/severely beaten] by her husband.” We randomized two elements of the story: whether the survivor was rich or poor and whether the survivor was beaten or severely beaten.

Our primary outcome measure is survivor and witness willingness to report IPV. First, enumerators asked, “How do you think the [rich/poor] woman should respond to being beaten/severely beaten by her husband?” Our primary pre-specified outcome coded responses as 1 if the respondent selected “report the incident to the police,” or “report the incident to the village chairperson,” and 0 otherwise. As we discuss in more detail below, we also report treatment effects on whether the respondent took any action (e.g., did not say “resolve the problem quietly with her husband”) and preferences for reporting to each local authority independently.

Enumerators then asked respondents about how they would respond to a situation if they had witnessed the incident: “Now I would like to ask how YOU would respond if you witnessed a [rich/poor] woman in your village who was [beaten/severely] by her husband. If YOU witnessed a [rich/poor] woman being [beaten/severely beaten], how would you respond?” Respondents were then provided the same list of answer options as the previous question. We coded responses as 1 if the respondent indicated that they would report the incident to the police or village chairperson, and 0 otherwise.

To elicit beliefs about the acceptability of intimate partner violence in gender, respondents asked, “In your opinion, does a man have good reason to hit his wife if she disobeys or disrespects him?” The answer options were “yes” or “no.” We coded responses as 1 if respondents did not believe a man has a good reason to hit his wife, and 0 otherwise.

Enumerators also asked respondents about their perceptions of their community’s likelihood of reporting IPV. Enumerators asked, “how many [rich/poor] women in your village do you think would report this situation to the local authority or police if they were [beaten/severely] by their husbands?” The answer options included: “None,” “A few,” “About half,” “Many,” and “All.” We coded the outcome as a continuous measure from “None” (coded 0) to “All” (coded 4).

To asses respondents’ beliefs about the likely consequences of an IPV survivor reporting their abuse, we read a second short vignette: “Imagine that a [rich/poor] woman in your village was [beaten/severely beaten] by her husband, so she chose to go to the [police/local authority] to report the incident.” Several characteristics were once again randomized: whether the survivor was rich or poor, whether the survivor was beaten or severely beaten, and whether the survivor had reported the incident to the police or local authorities.

Enumerators then asked several questions about how different actors would respond to the survivor’s report: whether the authority would doubt or question the story, whether the authority would take action to punish the perpetrator, whether the survivor’s in-laws would punish the survivor, whether the survivor’s husband would leave the survivor, and whether the community would support the survivor. For all questions, respondents could select from five options: not at all likely, unlikely, neither likely nor unlikely, somewhat likely, or very likely. For ease of interpretation, we recoded the direction of each outcome so that a higher value meant more support for (or less backlash to) the survivor. Lastly, we asked respondents an open-ended question, “Are there any worries a woman might have that might prevent her from reporting her husband?” and allowed respondents to provide an open-ended response.

2.6.2 Hypotheses

Before data collection, we pre-registered one-tailed tests for two groups of hypotheses (see Supplementary Material). First, we hypothesized that hearing the radio spot would increase willingness to say that survivors and witnesses of IPV should report IPV to authorities. Second, we hypothesized that hearing the radio spot would increase rejection of IPV, increase belief that others in the respondents’ community would report IPV, increase belief that authorities would believe and respond to survivors when they report IPV, increase belief that reporters’ family members will not take punitive action against survivors when they report IPV, and increase belief that reporters’ community members will support survivors when they report IPV.

2.7 Estimation

We estimate the effect of the radio spot using ordinary least squares regression. Let 𝑌_𝑖 denote the survey outcome for subject 𝑖, and let 𝑇_𝑖 denote this subject’s assigned treatment. The regression model:

$\mathit{Yi}=\mathit{\beta Ti}+\mathit{ui}$

Expresses the outcome as a linear function of the randomly assigned treatment and an unobserved disturbance term 𝑢_𝑖. The key parameter of interest is 𝛽, which represents the Average Treatment Effect. Unless otherwise specified, we will conduct one-tailed hypothesis tests in the direction specified by the pre-analysis plan.

3 Results

3.1 Descriptive results

3.1.1 Preferences for survivor and witness reporting of IPV

Before turning to experimental results, we present insights from our descriptive data on beliefs, attitudes, perceived norms, and behavioral intentions around reporting intimate partner violence. Table 2 shows the proportion of respondents in the control group who gave each of the possible answers to our two primary outcome measures. The results are divided between respondents who were randomly assigned to hear a vignette in which a man beats (kupiga) his wife or severely beats (kupiga vibaya sana) his wife.

Table 2

Table 2. Attitudes towards survivor and witness reporting of IPV.

Our first primary outcome is how a respondent would recommend that a survivor of IPV respond to being abused. In the standard case, respondents in the control group were roughly evenly divided between recommending three courses of action: resolving the problem quietly with the husband (29%), trying to work it out within the family (including parents or siblings) (24.5%), and reporting the incident to the village chairperson (27%). Fewer respondents would recommend asking for help from other community members (4%), religious leaders (6%), and the police (9%). In the case of significant abuse, respondents were 11 percentage points less likely to support quiet resolution with the husband. Instead, more respondents recommended working within the family (+4 pp), or reporting to police (+2 pp), and especially the village chairperson (+6.4 pp).

What about if a citizen witnessed an instance of IPV? When asked how they would respond to witnessing IPV, more than half of citizens said that they would either not get involved (23%) or speak only to the husband (36%). Just 29.5% said that they would report the abuse to the village chairperson and less than 1% said that they would report the abuse to the police. These numbers changed very little in a case of severe IPV: respondents were 6 percentage points less likely to avoid getting involved, preferring instead to speak directly to the husband (+4 pp), the village chairperson (+1.7 pp), or the police (+2.8 pp). The overall picture suggests a setting in which the majority of rural villagers would prefer to keep IPV “within the family,” and scarcely any villages would consider reporting IPV to formal law enforcement.

3.1.2 Beliefs about consequences of reporting IPV

Table 2 reveals a widespread preference for reporting IPV to local village chairpersons over local police. Two possible explanations are concerns about logistics and the likely reactions of officers: police stations are often harder to reach than the offices of local political representatives, and citizens may be more afraid of interactions with formal law enforcement.

To formally assess perceived differences in the consequences of reporting IPV to the village chair-person or the local police, we presented respondents with a vignette about an IPV survivor who decides to report her situation to a local authority, and randomized whether the authority was the village chairperson or the police. We then asked respondents about the likelihood of different outcomes of her decision: that the [village chairman/police] would doubt her story, that the [village chairman/police] would take action, that her in-laws would retaliate, that the survivor’s husband would leave her, and that the community would support her. All variables are coded from 0 to 3, where 3 means that the most positive outcome is considered more likely (or the negative outcome is considered unlikely).

Table 3 reveals stark differences in the perceived consequences of reporting to the police versus the village chairperson. On average, respondents were more likely to say that a hypothetical survivor who reported her abuse to the village chairman would be believed by authorities (coefficient = 0.19, p-value = 0.076) and would not be abandoned by her husband (coefficient = 0.21, p-value = 0.023) than a survivor who reported her abuse to the police. These findings are in keeping with the observations that village chairpersons are more familiar to the village and are more likely to handle problems discreetly. However, the results do not confirm other common anecdotal observations: reporting to the police is more likely to engender backlash from the woman’s in-laws or neighbors.

Table 3

Table 3. Beliefs about consequences of survivor reporting IPV to village chair or police.

Although respondents believe that the village chairperson is more likely to believe the survivor, they are much less likely to believe that the village chairpersons will take action to punish the abusive husband: 23% of the respondents said that it was “not at all likely” that the village chairman would punish the husband, relative to just 12% who said that police were not at all likely to do so (p value = 0.004).

In short, respondents view reporting IPV to the police as a more assertive response that is simultaneously more likely to lead to meaningful consequences for the perpetrator of abuse and more likely to engender backlash. It is important to note that these findings only represent villagers’ perceptions about the consequences of reporting. An important area for future research is to evaluate the validity of these beliefs.

3.2 Effect of radio drama on attitudes towards IPV reporting

3.2.1 Survivor reporting

Having documented widespread reluctance to report IPV, we now ask whether a short radio message designed to encourage reporting influences attitudes towards reporting. We begin our discussion of results with the most straightforward outcome measure: belief that survivors should report inti-mate partner violence. We first told a short story of a woman who was beaten by her husband. We then asked respondents what they would recommend the survivor do. Respondents were provided answer choices that were coded into three categories: taking any action at all to report, reporting the incident to the village chairperson, and reporting the incident to the police.

Table 4 shows that 76% of control group respondents said they would recommend reporting the incident to any of the reporting options, 41% said they would recommend reporting to local authorities, while only 10% said they would recommend the survivor report to the police.

Table 4

Table 4. Effect of radio drama on attitudes towards survivor reporting.

How did hearing a radio spot about intimate partner violence affect respondents’ beliefs about whether IPV survivors should report their abuse? Table 4 shows that respondents who heard the drama were 7.2 percentage points more likely to say that an IPV survivor should take any action, beyond resolving the problem quietly with her husband (“Any Response”) (one-tailed p-value = 0.002).

Respondents in the treatment group were also 4.7 percentage points more likely to recommend reporting the incident to either the police or the village chairperson (the primary outcome specified in our pre-analysis plan), although the effect falls short of statistical significance at the 0.05 level (one-tailed p-value = 0.067).

However, the measure combining all state authorities masks important variation: we observe a slightly negative effect of the drama on survivors’ preference for reporting to the village chairperson (coefficient = −4.2 pp., p-value = 0.932) but a substantial positive effect on willingness to report to the police (coefficient = 8.9 pp., p-value < 0.001). The drama moderately increased willingness to report to authorities but dramatically shifted beliefs about which authorities to report to.

3.2.2 Witness reporting

We also investigated whether the radio drama changed participants’ views about whether witnesses should report IPV. Using the same short story of a woman being beaten by her husband, enumerators asked participants how they would respond to witnessing this instance of intimate partner violence. Respondents were provided the same answer choices as the previous question, which were coded into three categories: taking any action at all to report, reporting the incident to local authorities (village chairperson or police), and specifically reporting to the police. It’s important to note that the radio drama did not model witnesses reporting, but rather modeled supporting a survivor who chose to report IPV. Reporting IPV without the consent of the survivor is not necessarily the best outcome in real-world settings.

Table 5 shows that while about 79% of control group respondents said they would report a witnessed incident of intimate partner violence in some way, only 32% said they would report the incident to local authorities, and only 2% of respondents said they would report a witnessed incident of IPV specifically to the police.

Table 5

Table 5. Effect of radio drama on attitudes towards witness reporting.

We do not observe statistically significant effects of the radio drama on respondents’ willingness to report IPV if they witness abuse. Treated respondents were slightly more likely to respond in some way (coefficient = 2.3 pp., one-tailed p-value = 0.181) or report to police (coefficient = 1.5 pp., one-tailed p-value = 0.083), but were slightly less likely to report witnessed abuse to the village chairperson (−5.4 pp., one-tailed p-value = 0.973).

Taken together, the preceding results suggest that the short radio spot generated a substantial but narrow effect on attitudes towards reporting intimate partner violence. Contrary to the primary hypothesis registered in the pre-analysis plan, we do not observe an aggregate effect of the drama on reporting behaviors Table A2. While the drama substantially increased the proportion of respondents who preferred the specific behavior modeled in the narrative (reporting IPV to the police) it had no discernible influence on, and may have traded off with, preferences for reporting IPV to local political leaders.

3.3 Effect of drama on drivers of IPV reporting

We now turn our attention to whether the drama influenced any of the three hypothesized obstacles to IPV reporting: acceptance of IPV, perceptions of community norms around IPV reporting, and beliefs about responses to IPV reporters by the survivors’ family, community, and state authorities.

3.3.1 Attitudes towards intimate partner violence

We begin our discussion of mechanisms with attitudes towards intimate partner violence. How did hearing a radio drama about intimate partner violence affect respondents’ views about the accept-ability of IPV? Table 6 shows that respondents who heard the radio spot were 2.2 percentage points more likely to say that a husband does not have a good reason to hit his wife if she disobeys him. However, the p-value of 0.229 falls short of conventional statistical significance levels.

Table 6

Table 6. Effect of radio drama on attitudes towards intimate partner.

3.3.2 Perceptions of IPV reporting norms

To measure respondents’ perceptions of community norms around IPV reporting, enumerators asked respondents what proportion of women in their village would report IPV to authorities if they were victimized. The estimated effect of the treatment on the perceived likelihood of survivor reporting (coefficient = 0.049, one-tailed p-value = 0.235) is small and is not statistically distinguishable from zero (see Table 7).

Table 7

Table 7. Effect of radio drama on perceptions of reporting norms.

3.3.3 Perceived responses to IPV reports

A final mechanism by which radio dramas might influence respondents’ willingness to report IPV is by changing respondents’ beliefs about the likely effects of the decision to report. Therefore the final set of outcome measures concerns respondents’ beliefs about responses to a survivor reporting IPV.

Table 8 offers little evidence that the drama influenced beliefs about the consequences of IPV reporting. The treatment increased respondents’ beliefs that authorities would believe a survivor who reports her abuser by just 0.036 on a 0–3 scale (one-tailed p-value = 0.312). However, Table A3 shows that the drama made respondents more likely to believe that police would believe a survivor while Table A4 shows no parallel effect on respondents’ beliefs about the village chairperson. This distinction may help explain why treatment increased respondents’ willingness to report to police but not village chairpersons.

Table 8

Table 8. Effect of radio drama on beliefs about effects of reporting.

The estimated effects of the drama on other beliefs about the consequences of survivor reporting are generally small and insignificant. The lone exception is that the drama made respondents more likely to believe that a survivor who reports to authorities would be punished by her in-laws, contrary to our prediction. The specter of familial backlash is raised in the drama, although the protagonist’s friends dismiss it. The combined results suggest that radio dramas do not cause citizens to substantially change their factual beliefs (the proportion of survivors who would report IPV in their community, responses to IPV reporting), even when they do shape respondents’ attitudes about appropriate actions.

3.4 Heterogeneous treatment effects

In line with our pre-analysis plan, we also evaluated whether the effect of treatment differed by several respondent characteristics: age, severity of abuse, wealth of survivor, religiosity, education, parental status, and media interest. There seems to be some variation in treatment effects on respondents’ willingness to report IPV to the police by respondents’ age, but there was no significant difference for any other reporting option. We found no significant heterogeneous effect based on respondents’ religiosity, education level, parental status, or interest in media. We also did not find statistically significant differences in responses based on the wealth of the survivor in the short story or the severity of the abuse mentioned in the short story. Overall, the impact of the radio message was surprisingly consistent across a variety of demographics.

4 Discussion

We set out to test the effect of a short, locally tailored radio spot on respondents’ willingness to report intimate partner violence. We hypothesized that a narrative drama modeling a woman’s decision to report her abuse would increase listeners’ support for reporting. Contrary to our expectations, we do not observe a substantively or statistically significant effect of the radio drama on our primary index measure, which combined survivor and witness willingness to report to police or local village chairpersons.

Instead, we find that the 1.5 min drama influenced only the specific behavior at the center of the narrative. Respondents were significantly more likely to endorse survivors reporting abuse to the police, but not more likely to endorse reporting to village chairpersons or witness reporting to the state. Nor did the drama influence related attitudes, perceived norms, or beliefs about the consequences of reporting intimate partner violence.

One potential concern is that these findings reflect respondents’ beliefs about the “appropriate answer” rather than meaningful changes in attitudes. However, null effects on a range of outcomes suggest that social desirability pressures alone cannot account for the results.

However, our findings should be interpreted in light of several limitations. The intervention was brief, so the findings may not generalize to longer or more multifaceted narrative campaigns. Because exposure and measurement occurred in a one-on-one survey setting, it remains unclear whether effects would persist over time or translate to more naturalistic listening environments. Finally, the study was conducted in rural northeastern Tanzania among low-income respondents with limited interaction with formal law enforcement. While such contexts are common in Sub-Saharan Africa, they represent important scope conditions for our conclusions.

Taken together, these findings contribute to the literatures on intimate partner violence reporting and entertainment-education in three ways. First, whereas much existing work evaluates sustained or episodic narrative interventions, we provide evidence on the effects—and limits—of short-form radio dramas commonly used in public service campaigns (Banerjee et al., 2019). Second, extending prior research that treats reporting as a unitary or binary outcome (Cooper et al., 2020; Green et al., 2020), we show that citizens sharply differentiate between reporting to formal law enforcement and reporting to local political authorities, with important implications for both barriers to and strategies for increasing IPV reporting. Third, by measuring multiple theorized mechanisms within a single experimental design, we offer a more direct test of how narrative dramas operate than many prior studies, which often infer mechanisms indirectly from observed behavioral change.

5 Conclusion

This study makes three contributions to the literature on entertainment-education and IPV prevention. First, we demonstrate that even brief narrative dramas commonly used in public health campaigns can influence behavioral intentions around IPV reporting. Second, we show that the effects of short narrative interventions are narrow: the drama influenced only the specific behavior it modeled—survivors reporting abuse to the police—with no spillover to related behaviors such as reporting to local leaders or witness reporting. These findings suggest that campaigns seeking broader change may require multiple, complementary messages that target distinct behaviors and audiences. Third, we highlight an underappreciated distinction in citizens’ perceptions of reporting to formal law enforcement versus local political authorities, clarifying how authority choice shapes responses to intimate partner violence in low-capacity settings. Together, these results underscore both the promise and the limits of narrative education approaches to IPV prevention.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by Lafayette College Institutional Review Board. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

KS: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing, Project administration. NR: Conceptualization, Methodology, Writing – review & editing, Data curation. KA: Conceptualization, Data curation, Writing – review & editing. DG: Conceptualization, Data curation, Writing – review & editing, Formal analysis, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft. BN: Conceptualization, Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by Lafayette College EXCEL Program.

Acknowledgments

The authors would like to thank Savannah Nivin, Benjamin Risley, Amber Peterman, Donald P Green, Alexander Brown, Isabella Tang, Athena Wagner, and participants at the Pennsylvania Political Science Conference and the Lafayette Undergraduate Research Conference for invaluable feed-back. We thank Andrew Clarke, eZakayo Zakaria, Marco Kitundu, Salma Emmanuel, Francis Nyonzo, and Constantine Manda for research support. We are indebted to the Economic and Social Research Foundation’s incredible field team, led ably by Neema Msechu and Andrew Lusakelo and overseen by Danford Sango: Husna Majura, Michael Zuakulu, Asha Matimu, David Lumato, Sylivana Karia, Anselem Kadallah, Shehe Omary, Angela Kitali, Hamza Mtinanga, Careen Ringo, Thomas Nkondola, and Vailet Urio.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcomm.2025.1707577/full#supplementary-material

References

Abramsky, T., Lees, S., Stöckl, H., Harvey, S., Kapinga, I., Ranganathan, M., et al. (2019). Women’s income and risk of intimate partner violence: secondary findings from the MAISHA cluster randomised trial in North-Western Tanzania. BMC Public Health 19:1108. doi: 10.1186/s12889-019-7454-1,

PubMed Abstract | Crossref Full Text | Google Scholar

Aiko, R. (2015). Effect of police integrity, government performance in fighting crime and accessi-bility of police stations on reporting of crime in Tanzania. Afrobarometer Policy Pap. 20, 1–23.

Google Scholar

Arias, E. (2019). How does media influence social norms? Experimental evidence on the role of common knowledge. Polit. Sci. Res. Methods 7, 561–578. doi: 10.1017/psrm.2018.1

Crossref Full Text | Google Scholar

Bandura, A. (2001). Social cognitive theory of mass communication. Media Psychol. 3, 265–299. doi: 10.1207/S1532785XMEP0303_03

Crossref Full Text | Google Scholar

Bandura, A. (2004). “Social cognitive theory for personal and social change by enabling media” in Entertainment-education and social change: History, research, and practice, LEA’S communication series (Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers), 75–96.

Google Scholar

Banerjee, A., Ferrara, E. L., and Orozco, V. (2019). Entertainment, education, and attitudes toward domestic violence. AEA Pap. Proc. 109, 133–137. doi: 10.1257/pandp.20191073

Crossref Full Text | Google Scholar

Bicchieri, C. (2016). Norms in the wild: How to diagnose, measure, and change social norms. Oxford, En: Oxford University Press.

Google Scholar

Bilali, R. (2022). Fighting violent extremism with narrative intervention: evidence from a field exper-iment in West Africa. Psychol. Sci. 33, 184–195.

Google Scholar

Blair, G., Littman, R., and Paluck, E. L. (2019). Motivating the adoption of new community-minded behaviors: an empirical test in Nigeria. Sci. Adv. 3, 1–8. doi: 10.1126/sciadv.aau5175

Crossref Full Text | Google Scholar

Blaydes, L., Fearon, J. D., and MacDonald, M. (2025). Understanding intimate partner violence. Annu. Rev. Polit. Sci. 28, 351–374. doi: 10.1146/annurev-polisci-041322-044022

Crossref Full Text | Google Scholar

Bourey, C., Williams, W., Bernstein, E. E., and Stephenson, R. (2015). Systematic review of struc-tural interventions for intimate partner violence in low- and middle-income countries: organizing evidence for prevention. BMC Public Health 15:1165. doi: 10.1186/s12889-015-2460-4,

PubMed Abstract | Crossref Full Text | Google Scholar

Boyer, C., Levy Paluck, E., Annan, J., Nevatia, T., Cooper, J., Namubiru, J., et al. (2022). Religious leaders can motivate men to cede power and reduce intimate partner vi-olence: experimental evidence from Uganda. Proc. Natl. Acad. Sci. 119:e2200262119. doi: 10.1073/pnas.2200262119

Crossref Full Text | Google Scholar

Bullock, O. M., Shulman, H. C., and Huskey, R. (2021). Narratives are persuasive because they are easier to understand: examining processing fluency as a mechanism of narrative persuasion. Front. Commun. 6, 1–12. doi: 10.3389/fcomm.2021.719615

Crossref Full Text | Google Scholar

Christopher, E., Drame, N. D., Leyna, G. H., Killewo, J., Bärnighausen, T., and Rohr, J. K. (2022). Dis-closure of intimate partner violence by men and women in Dar Es Salaam, Tanzania. Front. Public Health 10, 1–8. doi: 10.3389/fpubh.2022.928469

Crossref Full Text | Google Scholar

Clark, C. J., Ferguson, G., Subedi, S., Lad, A., Hadd, A. R., Shrestha, B., et al. (2024). Social norms, diffusion, and women’s risk of intimate partner violence in Nepal: impact assessment of a social and behavior change communication intervention (change starts at home). SSM Popul. Health 25:101583. doi: 10.1016/j.ssmph.2023.101583,

PubMed Abstract | Crossref Full Text | Google Scholar

Cooper, J., Green, D. P., and Wilke, A. M. (2020). Reducing violence against women in Uganda through video dramas: a survey experiment to illuminate causal mechanisms. AEA Pap. Proc. 110, 615–619. doi: 10.1257/pandp.20201048

Crossref Full Text | Google Scholar

del Eggers Campo, I., and Steinert, J. I. (2022). The effect of female economic empowerment interven-tions on the risk of intimate partner violence: a systematic review and meta-analysis. Trauma Violence Abuse 23, 810–826. doi: 10.1177/1524838020976088

Crossref Full Text | Google Scholar

Dickson, K. S., Boateng, E. N. K., Adzrago, D., Addo, I. Y., Acquah, E., and Nyarko, S. H. (2023). Silent suffering: unveiling factors associated with women’s inability to seek help for intimate partner violence in sub-Saharan Africa (SSA). Reprod. Health 20:110. doi: 10.1186/s12978-023-01651-7,

PubMed Abstract | Crossref Full Text | Google Scholar

Dillon, G., Hussain, R., Loxton, D., and Rahman, S. (2013). Mental and physical health and inti-mate partner violence against women: a review of the literature. Int. J. Fam. Med. 2013:313909. doi: 10.1155/2013/313909,

PubMed Abstract | Crossref Full Text | Google Scholar

Doré, A., Rebitschek, F. G., Morello, K., Stierstädter, E., Thiemann, K. J., and Kubiak, T. (2025). A systematic review and meta-analysis of the effects of entertainment-education interventions on persuasive health outcomes. Sci. Rep. 15:27939. doi: 10.1038/s41598-025-11012-y,

PubMed Abstract | Crossref Full Text | Google Scholar

Ellsberg, M., Arango, D. J., Morton, M., Gennari, F., Kiplesund, S., Contreras, M., et al. (2015). Prevention of violence against women and girls: what does the evidence say? Lancet 385, 1555–1566. doi: 10.1016/S0140-6736(14)61703-7,

PubMed Abstract | Crossref Full Text | Google Scholar

Felson, R. B., Messner, S. F., and Hoskin, A. (1999). The victim-offender relation-ship and calling the police in assaults. Criminology 37, 931–948. doi: 10.1111/j.1745-9125.1999.tb00510.x

Crossref Full Text | Google Scholar

Felson, R. B., and Paré, P.-P. (2005). The reporting of domestic violence and sexual assault by non-strangers to the police. J. Marriage Fam. 67, 597–610. doi: 10.1371/journal.pdig.0000358

Crossref Full Text | Google Scholar

Ferraz, C., and Schiavon, L. 2022 Crime, punishment, and prevention: the effect of a legal reform on violence against women. Rochester NY: Social Science Research Network.

Google Scholar

Filippo, A. D., Bellatin, P., Tietz, N., Grant, E., Whitefield, A., Nkopane, P., et al. (2023). Effects of digital chatbot on gender attitudes and exposure to intimate partner violence among young women in South Africa. PLoS Digit. Health 2:e0000358.

Google Scholar

Fotini, C., Larreguy, H., Parker-Magyar, E., and Quintero, M. (2023). Empowering women fac-ing gender-based violence amid COVID-19 through media campaigns. Nat. Hum. Behav. 7, 1740–1752. doi: 10.1038/s41562-023-01665-y

Crossref Full Text | Google Scholar

Green, D. P., Groves, D. W., and Manda, C. (2021). A radio Drama’s effects on HIV attitudes and policy priorities: a field experiment in Tanzania. Health Educ. Behav. 48, 842–851. doi: 10.1177/10901981211010421,

PubMed Abstract | Crossref Full Text | Google Scholar

Green, D. P., Groves, D. W., Manda, C., Montano, B., and Rahmani, B. (2023). A radio drama’s effects on attitudes toward early and forced marriage: results from a field experiment in rural Tanzania. Comp. Polit. Stud. 56, 1115–1155. doi: 10.1177/00104140221139385

Crossref Full Text | Google Scholar

Green, D. P., Groves, D. W., Manda, C., Montano, B., and Rahmani, B. (2024). The effects of indepen-dent local radio on Tanzanian public opinion: evidence from a planned natural experiment. J. Polit. 86, 231–240. doi: 10.1086/726964

Crossref Full Text | Google Scholar

Green, D. P., Wilke, A. M., and Cooper, J. (2020). Countering violence against women by encouraging disclosure: a mass media experiment in rural Uganda. Comp. Polit. Stud. 53, 2283–2320. doi: 10.1177/0010414020912275

Crossref Full Text | Google Scholar

Groves, D. W. (2023). Essays on Media and Accountability. New York City, NY: Columbia University.

Google Scholar

Haushofer, J., Ringdal, C., Shapiro, J. P., and Wang, X. Y. 2019 Income Changes and Intimate Partner Violence: Evidence from Unconditional Cash Transfers in Kenya. National Bureau of Economic Research.

Google Scholar

Heise, L. (1998). Violence against women: an integrated, ecological framework. Violence Against Women 4, 262–290. doi: 10.1177/1077801298004003002

Crossref Full Text | Google Scholar

Hoff, K., Jalan, J., and Santra, S. 2021 Participatory theater empowers women: evidence from India. World Pank Policy Research Working Paper.

Google Scholar

Jassal, N. (2020). Gender, law enforcement, and access to justice: evidence from all-women police stations in India. Am. Polit. Sci. Rev. 114, 1035–1054. doi: 10.1017/S0003055420000684

Crossref Full Text | Google Scholar

Jassal, N. (2024). Does victim gender matter for justice delivery? Police and judicial responses to women’s cases in India. Am. Polit. Sci. Rev. 118, 1278–1304. doi: 10.1017/S0003055423000916

Crossref Full Text | Google Scholar

Kapiga, S., Harvey, S., Mshana, G., Hansen, C. H., Mtolela, G. J., Madaha, F., et al. (2019). A social empowerment intervention to prevent intimate partner violence against women in a microfinance scheme in Tanzania: findings from the MAISHA cluster randomised controlled trial. Lancet Glob. Health 7, e1423–e1434. doi: 10.1016/S2214-109X(19)30316-X

Crossref Full Text | Google Scholar

Katiti, V., Sigalla, G. N., Rogathi, J., Manongi, R., and Mushi, D. (2016). Factors influencing disclosure among women experiencing intimate partner violence during pregnancy in Moshi municipality, Tanzania. BMC Public Health 16:715. doi: 10.1186/s12889-016-3345-x

Crossref Full Text | Google Scholar

Krause, A. E., and Fletcher, H. (2025). Radio relationships: perceptions on listenership and well-being. Australas. J. Ageing 44:e70008. doi: 10.1111/ajag.70008,

PubMed Abstract | Crossref Full Text | Google Scholar

Lappeman, J., Broster, P., Madurai, M., Masipa, M. S., and Cheteni, P. (2025). “Can we be young and carefree? Using the Roy adaptation model to explore youth pressure in South Africa” in Young Consumers: Insight and Ideas for Responsible Marketers, 1–20. Emerald Publishing.

Google Scholar

Lazarev, E. (2023). State-building as Lawfare: Custom, sharia, and state law in postwar Chechnya. Cambridge, En: Cambridge University Press.

Google Scholar

Lees, S., Marchant, M., and Desmond, N. (2021). Addressing intimate partner violence using gender-transformative approaches at a community level in rural Tanzania: the UZIKWASA program. J. Interpers. Violence 36, NP7791–NP7812. doi: 10.1177/0886260519831373

Crossref Full Text | Google Scholar

McCleary-Sills, J., Namy, S., Nyoni, J., Rweyemamu, D., Salvatory, A., and Steven, E. (2016). Stigma, shame and women’s limited agency in help-seeking for intimate part-ner violence. Glob. Public Health 11, 224–235. doi: 10.1080/17441692.2015.1047391

Crossref Full Text | Google Scholar

Montano, B. 2025. Misperceived gender norms and women’s strategic political advocacy: experi-mental evidence from rural Tanzania. Working Paper.

Google Scholar

Montano, B., Emmanuel, S., Green, D. P., Groves, D. W., and Rahmani, B. 2023. Can a Radio Drama Influence Awareness, Prioritization, and Willingness to Respond to Gender Based Violence? Evidence from a Field Experiment in Tanzania. Working Paper.

Google Scholar

Mootz, J. J., Spencer, C. M., Ettelbrick, J., Kann, B., dos Fortunato Santos, P., Palmer, M., et al. (2023). Risk markers for victimization and perpetration of male-to-female physical intimate partner violence in sub-Saharan Africa: a meta-analysis. Trauma Violence Abuse 24, 3433–3444. doi: 10.1177/15248380221129589

Crossref Full Text | Google Scholar

Mossie, T. B., Mekonnen Fenta, H., Tadesse, M., and Tadele, A. (2023). Mapping the disparities in intimate partner violence prevalence and determinants across sub-Saharan Africa. Front. Public Health 11:1188718. doi: 10.3389/fpubh.2023.1188718,

PubMed Abstract | Crossref Full Text | Google Scholar

Murrar, S., and Brauer, M. (2018). Entertainment-education effectively reduces prejudice. Group Process. Intergroup Relat. 21, 1053–1077. doi: 10.1177/1368430216682350

Crossref Full Text | Google Scholar

Nabaggala, M. S., Reddy, T., and Manda, S. (2021). Effects of rural–urban residence and education on intimate partner violence among women in sub-Saharan Africa: a meta-analysis of health survey data. BMC Womens Health 21:149. doi: 10.1186/s12905-021-01286-5,

PubMed Abstract | Crossref Full Text | Google Scholar

Nasir, A., and Naaz, A. (2024). The role of religion in shaping cultural norms. Applied Ling. Res. Rev. 1, 32–40.

Google Scholar

Okenwa, L. E. E., Lawoko, S., and Jansson, B. (2009). Factors associated with disclosure of intimate partner violence among women in Lagos, Nigeria. J. Injury Viol. Res. 1, 37–47. doi: 10.5249/jivr.v1i1.15,

PubMed Abstract | Crossref Full Text | Google Scholar

Palermo, T., Bleck, J., and Peterman, A. (2014). Tip of the iceberg: reporting and gender-based violence in developing countries. Am. J. Epidemiol. 179, 602–612. doi: 10.1093/aje/kwt295,

PubMed Abstract | Crossref Full Text | Google Scholar

Paluck, E. L., and Green, D. P. (2009). Deference, dissent, and dispute resolution: an experimental intervention using mass media to change norms and behavior in Rwanda. Am. Polit. Sci. Rev. 103, 622–644. doi: 10.1017/S0003055409990128

Crossref Full Text | Google Scholar

Peterman, A. (2025). Edutainment to prevent violence against women and children. World Bank Res. Obs. :lkaf002. doi: 10.1093/wbro/lkaf002

Crossref Full Text | Google Scholar

Rahmani, B. (2025a). Essays on Media and Political Persuasion in Autocracies. New York City, NY: Columbia University.

Google Scholar

Rahmani, B. (2025b). How autocrats use proxies to control the media. J. Democr. 36, 92–105. doi: 10.1353/jod.2025.a954566

Crossref Full Text | Google Scholar

Rahmani, B., Green, D. P., Groves, D. W., Emmanuel, S., and Montano, B.. (2024). Radio dramas can build political support for environmental protection: experimental evidence from rural Tanzania. Working Paper.

Google Scholar

Rahmani, B., Montano, B., Groves, D. W., and Green, D. P. (2025). The persuasive effects of narrative entertainment: a meta-analysis of recent experiments. Behav. Public Policy, 1–23. doi: 10.1017/bpp.2025.10010

Crossref Full Text | Google Scholar

Reese, B. M., Chen, M. S., Nekkanti, M., and Mulawa, M. I. (2021). Prevalence and risk factors of women’s past-year physical IPV perpetration and victimization in Tanzania. J. Interpers. Violence 36, 1141–1167. doi: 10.1177/0886260517738775

Crossref Full Text | Google Scholar

Rohn, E., and Tenkorang, E. Y. (2023). Structural and institutional barriers to help-seeking among female victims of intimate partner violence in Ghana. J. Fam. Violence 38, 815–827. doi: 10.1007/s10896-022-00433-2

Crossref Full Text | Google Scholar

Sanin, D. (2021). Do domestic violence laws protect women from domestic violence? evidence from Rwanda. Washington DC: Georgetown University Initiative on Innovation, Development and Evaluation.

Google Scholar

Simmons, E., Halim, N., Servidone, M., Steven, E., Reich, N., Badi, L., et al. (2020). Prevention and mitigation of intimate-partner violence: the role of community leaders in Tanzania. Violence Against Women 26, 359–378. doi: 10.1177/1077801219832923,

PubMed Abstract | Crossref Full Text | Google Scholar

Slater, M. D., and Rouner, D. (2002). Entertainment-education and elaboration likelihood: understand-ing the processing of narrative persuasion. Commun. Theory 12, 173–191. doi: 10.1093/ct/12.2.173

Crossref Full Text | Google Scholar

Sturmer, M. (1998). The media history of Tanzania. Salzburg, Germany: Ndanda Mission Press.

Google Scholar

Tankard, M. E., and Paluck, E. L. (2016). Norm perception as a vehicle for social change. Soc. Issues Policy Rev. 10, 181–211. doi: 10.1111/sipr.12022

Crossref Full Text | Google Scholar

Tanzania Ministry of Health (2023). Tanzania Demographic and Health Survey and Malaria Indicator Survey 2022 Final Report. DHS Final Report FR382. Dodoma, Tanzania, and Rockville, Maryland: Ministry of Health (MoH) [Tanzania Mainland & Zanzibar]; National Bureau of Statistics (NBS); Office of the Chief Government Statistician (OCGS); and ICF/The DHS Program.

Google Scholar

Umubyeyi, A., Persson, M., Mogren, I., and Krantz, G. (2016). Gender inequality prevents abused women from seeking care despite protection given in gender-based violence legislation: a qual-itative study from Rwanda. PLoS One 11:e0154540 Publisher: Public Library of Science. doi: 10.1371/journal.pone.0154540,

PubMed Abstract | Crossref Full Text | Google Scholar

United Republic of Tanzania (2024). National plan of action to end violence against women and children (NPA – VAWC II 2024/25–2028/29). Dodoma, Tanzania: National Action Plan, Ministry of Community Development, Gender, Women and Special Groups.

Google Scholar

Vyas, S., Meinhart, M., Troy, K., Brumbaum, H., Poulton, C., and Stark, L. (2023). The economic cost of violence against women and girls in low- and middle-income countries: a systematic review of the evidence. Trauma Violence Abuse 24, 44–55. doi: 10.1177/15248380211016018,

PubMed Abstract | Crossref Full Text | Google Scholar

Wang, H., and Singhal, A. (2021). “Theorizing entertainment-education: a complementary perspective to the development of entertainment theory” in The Oxford handbook of entertainment theory (New York, NY: Oxford University Press), 819–838.

Google Scholar

World Health Organization (2023). Violence against women prevalence estimates, 2023: global, regional and national prevalence estimates for intimate partner violence against women and non-partner sexual violence against women. Geneva: World Health Organization; 2025. Licence: CC BY-NC-SA 3.0 IGO

Google Scholar

World Bank. March (2022). Tanzania GBV Assessment. © World Bank.

Google Scholar

Yonga, A. M., Kiss, L., and Onarheim, K. H. (2022). A systematic review of the effects of intimate part-ner violence on HIV-positive pregnant women in sub-Saharan Africa. BMC Public Health 22:220. doi: 10.1186/s12889-022-12619-w,

PubMed Abstract | Crossref Full Text | Google Scholar

Yount, K. M., Cheong, Y. F., Bergenfeld, I., Trang, Q. T., Sales, J. M., Li, Y., et al. (2023). Im-pacts of GlobalConsent, a web-based social norms edutainment program, on sexually violent be-havior and bystander behavior among university men in Vietnam: randomized controlled trial. JMIR Public Health Surveill. 9:e35116. doi: 10.2196/35116,

PubMed Abstract | Crossref Full Text | Google Scholar