The Entertaining Way to Behavioral Change: Fighting HIV with MTV
Summary:
Nigeria has a generalized HIV epidemic, which requires scalable and low-cost interventions targeting the general population. This study evaluated the effectiveness of MTV Shuga in Nigeria with 18-25-year-olds from the general population. Using a cluster randomized control trial, this trial measured Shuga's effectiveness on HIV outcomes eight months after program exposure. The trial was designed to also measures program spillover effects on viewers' friends; and the mediating effects of social norms, program immersion and character-identification. The study found statistical and clinical effects on viewers' increased knowledge about HIV transmission (0.13 standard deviations) and improved attitudes towards HIV+ people (0.10 SDs). It found significant effects on objective measures for HIV testing uptake (doubling testing rates) and Chlamydia infections (halving infections for women). Men became half as likely to report concurrent sexual partners. We find evidence of knowledge spillovers on viewers' friends, though not on their attitudes and behaviors. While we find no evidence that social norms mediated program effects for HIV private behaviors, we find that effects were larger for viewers that experienced greater program immersion and greater character-identification. These storyteller mediators are the comparative advantage of edu-tainment narratives over traditional educational approaches. In addition to strengthening the business case that edu-tainment is a smart investment, the trial's methodological innovations show that media broadcasts and its mediators can be rigorously measured.
Background/Objectives:
Through a field experiment we evaluate the impact of Shuga 8 months after program exposure. Through self-reported and objective measures, we answer the following effectiveness questions: (1) What is the program's impacts on HIV knowledge and attitudes; on HIV testing uptake and risky sexual behaviors? (2) Does it have spillover effects on viewers' friends? The study design also allows testing potential mediators (3) Do impacts on social norms mediate program impacts? And (4) Are effects stronger for viewers who experience greater program immersion and greater character-identification? These four questions are major contributions to the edu-tainment literature.
Description of Intervention and/or Methods/Design:
18-25-year-old youth were invited to a series of community entertainment screenings in 80 communities. The control was exposed to a placebo TV drama to net-out the delivery mechanism of community screenings. Surveys were collected at baseline and eight months after program exposure (n=4,986). Objective measures were collected after the post-intervention survey. Two experimental arms tested the role of social norms as potential mediators: (i) half of treatment community screenings were exposed to HIV views of local peers; and (ii) in a cross-cutting arm, half of individuals were given an extra ticket to bring a friend. To shed light on the workings of edutainment, the follow-up survey asked questions related to program immersion and character-identification. To test for potential spillovers, key for assessing community-level effects, we surveyed friends of main participants in both treatment and control groups.
Results/Lessons Learned:
Shuga increased viewers' knowledge about HIV transmission (0.13 standard deviations) and improved their attitudes towards HIV+ people (0.10 SDs). It doubled HIV testing rates. Men became half as likely to report concurrent sexual partners; and Chlamydia infections were halved for women (a similar effect was found for men, though this was not statistically significant). No effects were found on increased condom use. While we find no evidence that impacts on perceptions of social norms played a role for this private behavior (the literature founds effects for public behaviors), the study provides strong evidence about storytelling mediators. Effects were larger for viewers who experienced stronger program immersion; and greater character-identification. Moreover, while we find evidence that viewers' friends became more knowledgeable about HIV transmission, we found no effects on their HIV attitudes and behaviors. This suggests living the drama may be required for impacting these deep-seated outcomes.
Discussion/Implications for the Field:
The measurement of causal effects of media broadcasts remains rare in the literature, affecting its business case for development investments. Rigorously measuring mediators can improve their content design. The trial shows that field experiments with placebo controls help obtain reliable measures of media effectiveness; and that mediators can be measured when incorporated in study design. The Shuga trial provides strong evidence that edu-tainment narratives can be scalable and cost-effective intervention against the HIV epidemic. With the trial's results, the World Bank is currently conducting a cost-benefit analysis. Preliminary results suggest returns of $150 for every dollar invested.
Abstract submitted by:
Victor Orozco - World Bank
Abhijit Banerjee - Massachusettes Institute of Technology (MIT)
Eliana La Ferrara - Universita' Bocconi
Nigeria has a generalized HIV epidemic, which requires scalable and low-cost interventions targeting the general population. This study evaluated the effectiveness of MTV Shuga in Nigeria with 18-25-year-olds from the general population. Using a cluster randomized control trial, this trial measured Shuga's effectiveness on HIV outcomes eight months after program exposure. The trial was designed to also measures program spillover effects on viewers' friends; and the mediating effects of social norms, program immersion and character-identification. The study found statistical and clinical effects on viewers' increased knowledge about HIV transmission (0.13 standard deviations) and improved attitudes towards HIV+ people (0.10 SDs). It found significant effects on objective measures for HIV testing uptake (doubling testing rates) and Chlamydia infections (halving infections for women). Men became half as likely to report concurrent sexual partners. We find evidence of knowledge spillovers on viewers' friends, though not on their attitudes and behaviors. While we find no evidence that social norms mediated program effects for HIV private behaviors, we find that effects were larger for viewers that experienced greater program immersion and greater character-identification. These storyteller mediators are the comparative advantage of edu-tainment narratives over traditional educational approaches. In addition to strengthening the business case that edu-tainment is a smart investment, the trial's methodological innovations show that media broadcasts and its mediators can be rigorously measured.
Background/Objectives:
Through a field experiment we evaluate the impact of Shuga 8 months after program exposure. Through self-reported and objective measures, we answer the following effectiveness questions: (1) What is the program's impacts on HIV knowledge and attitudes; on HIV testing uptake and risky sexual behaviors? (2) Does it have spillover effects on viewers' friends? The study design also allows testing potential mediators (3) Do impacts on social norms mediate program impacts? And (4) Are effects stronger for viewers who experience greater program immersion and greater character-identification? These four questions are major contributions to the edu-tainment literature.
Description of Intervention and/or Methods/Design:
18-25-year-old youth were invited to a series of community entertainment screenings in 80 communities. The control was exposed to a placebo TV drama to net-out the delivery mechanism of community screenings. Surveys were collected at baseline and eight months after program exposure (n=4,986). Objective measures were collected after the post-intervention survey. Two experimental arms tested the role of social norms as potential mediators: (i) half of treatment community screenings were exposed to HIV views of local peers; and (ii) in a cross-cutting arm, half of individuals were given an extra ticket to bring a friend. To shed light on the workings of edutainment, the follow-up survey asked questions related to program immersion and character-identification. To test for potential spillovers, key for assessing community-level effects, we surveyed friends of main participants in both treatment and control groups.
Results/Lessons Learned:
Shuga increased viewers' knowledge about HIV transmission (0.13 standard deviations) and improved their attitudes towards HIV+ people (0.10 SDs). It doubled HIV testing rates. Men became half as likely to report concurrent sexual partners; and Chlamydia infections were halved for women (a similar effect was found for men, though this was not statistically significant). No effects were found on increased condom use. While we find no evidence that impacts on perceptions of social norms played a role for this private behavior (the literature founds effects for public behaviors), the study provides strong evidence about storytelling mediators. Effects were larger for viewers who experienced stronger program immersion; and greater character-identification. Moreover, while we find evidence that viewers' friends became more knowledgeable about HIV transmission, we found no effects on their HIV attitudes and behaviors. This suggests living the drama may be required for impacting these deep-seated outcomes.
Discussion/Implications for the Field:
The measurement of causal effects of media broadcasts remains rare in the literature, affecting its business case for development investments. Rigorously measuring mediators can improve their content design. The trial shows that field experiments with placebo controls help obtain reliable measures of media effectiveness; and that mediators can be measured when incorporated in study design. The Shuga trial provides strong evidence that edu-tainment narratives can be scalable and cost-effective intervention against the HIV epidemic. With the trial's results, the World Bank is currently conducting a cost-benefit analysis. Preliminary results suggest returns of $150 for every dollar invested.
Abstract submitted by:
Victor Orozco - World Bank
Abhijit Banerjee - Massachusettes Institute of Technology (MIT)
Eliana La Ferrara - Universita' Bocconi
Source
Approved abstract for the postponed 2020 SBCC Summit in Marrakech, Morocco. Provided by the International Steering Committee for the Summit. Image credit: Black Youth Project