Björn Rohr
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Björn Rohr has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
Henning Silber
3 shared publications- Barbara Felderer
Leibniz Institute for the Social Sciences, Mannheim, Germany
2 shared publications - Henning Silber
University of Michigan
2 shared publications - Jan Priebe
Affiliations Research Group Health Economics, Bernhard Nocht Institute for Tropical Medicine (BNITM), Hamburg, Germany, Hamburg Center for Health Economics (HCHE), University of Hamburg, Hamburg, Germany
2 shared publications - Annika Stein
GESIS Leibniz Institute for the Social Sciences
1 shared publication - Arne Bethmann
SHARE Berlin Institute, Germany
1 shared publication
Latest Publications
- Open Access
- 25 June 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e68472
Incentives and Trust Are the Main Drivers of Recruiting Participants in 6 African Countries via Web-Based Environments: A Vignette Survey Experiment
Background: In-person health surveys and biomarker collections (eg, blood testing) provide crucial data to monitor and investigate progress on health outcomes in sub-Saharan Africa. Bearing in mind that administrative sampling frames are often outdated and financial resources can be limited, it is of substantial policy importance to better understand whether recruitment of individuals for in-person health data collection efforts can be accomplished via web-based environments such as social media sites. Yet, there is little methodological research on (1) the feasibility of recruitment through web-based environments and (2) the factors that drive in-person survey participation rates in sub-Saharan Africa countries. Objective: This study aimed to share our experimental results on the recruitment of individuals from sub-Saharan Africa for participation in in-person, health-related surveys and biomarker collections via Facebook ads and to provide recommendations for future data collections and research. Methods: We conducted a preregistered 2×4×4 vignette experiment to investigate people’s willingness to participate in in-person health surveys and blood tests. The experiment was part of a web survey (n≈10,600) of individuals recruited via Facebook advertisements that we conducted in early 2023 in 6 sub-Saharan Africa countries (Ghana, Kenya, Nigeria, South Africa, Tanzania, and Uganda). Based on the theories of contextual integrity, economic participation, and social exchange, three factors were varied: (1) the topic (HIV or diabetes), (2) the incentive (US $0, US $2 cash, US $2 voucher, and US $2 lottery), and (3) the sponsor (nongovernmental organization, statistical office, health ministry, or local university). Results: Overall, we found that a majority of survey participants are willing to participate in in-person health surveys and provide biomarkers (vignette means range between 5.54 and 6.09 on a 1 to 7 scale). First, providing a financial incentive significantly increased the likelihood of being willing to participate (b=.180, .188, and .200; all P<.001). Second, individuals with high levels of trust in nongovernmental organizations or the health ministry were more likely to be willing to participate (b=.086 and .048; both P<.001). In contrast, 2 factors (topic and sponsor) showed mainly non-significant effects (b=.010, P=.63; b=.041, P=.18; b=.042, P=.19; b=.063, P=.05). Other factors that were related to an increase in willingness to participate included fertility levels (having children), risk-taking, having an illness (HIV, diabetes), better general health, social trust, trust in science, survey enjoyment, survey value, and cognitive skills. Conclusions: Together, the study’s results suggest that using a web-based environment for recruiting health research participants in sub-Saharan Africa can be a viable option and emphasize the importance of adequate compensation and trust in the sponsor. The findings also indicated that several attitudinal but almost none of the sociodemographic variables are systematically related to the willingness to participate in health-related in-person data collection activities.
- 25 June 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e68472
Open AccessIncentives and Trust Are the Main Drivers of Recruiting Participants in 6 African Countries via Web-Based Environments: A Vignette Survey Experiment
Background: In-person health surveys and biomarker collections (eg, blood testing) provide crucial data to monitor and investigate progress on health outcomes in sub-Saharan Africa. Bearing in mind that administrative sampling frames are often outdated and financial resources can be limited, it is of substantial policy importance to better understand whether recruitment of individuals for in-person health data collection efforts can be accomplished via web-based environments such as social media sites. Yet, there is little methodological research on (1) the feasibility of recruitment through web-based environments and (2) the factors that drive in-person survey participation rates in sub-Saharan Africa countries. Objective: This study aimed to share our experimental results on the recruitment of individuals from sub-Saharan Africa for participation in in-person, health-related surveys and biomarker collections via Facebook ads and to provide recommendations for future data collections and research. Methods: We conducted a preregistered 2×4×4 vignette experiment to investigate people’s willingness to participate in in-person health surveys and blood tests. The experiment was part of a web survey (n≈10,600) of individuals recruited via Facebook advertisements that we conducted in early 2023 in 6 sub-Saharan Africa countries (Ghana, Kenya, Nigeria, South Africa, Tanzania, and Uganda). Based on the theories of contextual integrity, economic participation, and social exchange, three factors were varied: (1) the topic (HIV or diabetes), (2) the incentive (US $0, US $2 cash, US $2 voucher, and US $2 lottery), and (3) the sponsor (nongovernmental organization, statistical office, health ministry, or local university). Results: Overall, we found that a majority of survey participants are willing to participate in in-person health surveys and provide biomarkers (vignette means range between 5.54 and 6.09 on a 1 to 7 scale). First, providing a financial incentive significantly increased the likelihood of being willing to participate (b=.180, .188, and .200; all P<.001). Second, individuals with high levels of trust in nongovernmental organizations or the health ministry were more likely to be willing to participate (b=.086 and .048; both P<.001). In contrast, 2 factors (topic and sponsor) showed mainly non-significant effects (b=.010, P=.63; b=.041, P=.18; b=.042, P=.19; b=.063, P=.05). Other factors that were related to an increase in willingness to participate included fertility levels (having children), risk-taking, having an illness (HIV, diabetes), better general health, social trust, trust in science, survey enjoyment, survey value, and cognitive skills. Conclusions: Together, the study’s results suggest that using a web-based environment for recruiting health research participants in sub-Saharan Africa can be a viable option and emphasize the importance of adequate compensation and trust in the sponsor. The findings also indicated that several attitudinal but almost none of the sociodemographic variables are systematically related to the willingness to participate in health-related in-person data collection activities.
- Open Access
- 11 June 2025
- Published by Taylor & Francis in International Journal of Social Research Methodology
- Vol. 29 (2) , 197-214
Education bias in probability-based surveys in Germany: evidence and possible solutions
- 11 June 2025
- Published by Taylor & Francis in International Journal of Social Research Methodology
- Vol. 29 (2) , 197-214
Open AccessEducation bias in probability-based surveys in Germany: evidence and possible solutions
- Open Access
- 19 February 2025
- Published by Center for Open Science in SocArXiv
Sampling for a Cross-National Survey in Six African Countries Using Social Media Advertisements: Comparison of Different Targeting Strategies, Types of Estimates, and Selectivity against Population Benchmarks
When researchers recruit participants for an online survey via Facebook advertisements, a targeting algorithm determines which individuals receive the survey invitation. Because the algorithm maximizes the number of clicks on the survey link, predominantly easy-to-reach population groups in a country or region might be reached (simple demographic targeting: SDT). To reduce representation bias, researchers can target selected demographic population groups separately (complex demographic targeting: CDT). In this study, we evaluate whether CDT effectively leads to less bias in univariate, bivariate, and multivariate estimates than SDT. This is done by comparing socio-demographic and health-related measures of health surveys recruited via Facebook with and without CDT in six African countries against probability-based survey benchmarks. Independent of the targeting strategy, our results show that many estimates were strongly biased, especially the univariate estimates of education and Internet use, whereas other estimates (e.g., HIV knowledge) were less affected. Although we found minor evidence that the CDT method reduced bias in univariate estimates, bias in relationship estimates stayed similar. Moreover, the study shows that univariate estimates are generally more often biased than relationship estimates and that bias is partially related to coverage issues. The success of weighting in bias reduction was mixed.
- 19 February 2025
- Published by Center for Open Science in SocArXiv
Open AccessSampling for a Cross-National Survey in Six African Countries Using Social Media Advertisements: Comparison of Different Targeting Strategies, Types of Estimates, and Selectivity against Population Benchmarks
When researchers recruit participants for an online survey via Facebook advertisements, a targeting algorithm determines which individuals receive the survey invitation. Because the algorithm maximizes the number of clicks on the survey link, predominantly easy-to-reach population groups in a country or region might be reached (simple demographic targeting: SDT). To reduce representation bias, researchers can target selected demographic population groups separately (complex demographic targeting: CDT). In this study, we evaluate whether CDT effectively leads to less bias in univariate, bivariate, and multivariate estimates than SDT. This is done by comparing socio-demographic and health-related measures of health surveys recruited via Facebook with and without CDT in six African countries against probability-based survey benchmarks. Independent of the targeting strategy, our results show that many estimates were strongly biased, especially the univariate estimates of education and Internet use, whereas other estimates (e.g., HIV knowledge) were less affected. Although we found minor evidence that the CDT method reduced bias in univariate estimates, bias in relationship estimates stayed similar. Moreover, the study shows that univariate estimates are generally more often biased than relationship estimates and that bias is partially related to coverage issues. The success of weighting in bias reduction was mixed.
- Open Access
- 28 September 2024
- Published by SAGE Publications in Sociological Methodology
- Vol. 55 (1) , 121-154
Comparing the Accuracy of Univariate, Bivariate, and Multivariate Estimates across Probability and Nonprobability Surveys with Population Benchmarks
Previous studies have shown many instances where nonprobability surveys were not as accurate as probability surveys. However, because of their cost advantages, nonprobability surveys are widely used, and there is much debate over the appropriate settings for their use. To contribute to this debate, we evaluate the accuracy of nonprobability surveys by investigating the common claim that estimates of relationships are more robust to sample bias than means or proportions. We compare demographic, attitudinal, and behavioral variables across eight German probability and nonprobability surveys with demographic and political benchmarks from the microcensus and a high-quality, face-to-face survey. In the analyses, we compare three types of statistical inference: univariate estimates, bivariate Pearson’s r coefficients, and 24 different multiple regression models. The results indicate that in univariate comparisons, nonprobability surveys were clearly less accurate than probability surveys when compared with the population benchmarks. These differences in accuracy were smaller in the bivariate and the multivariate comparisons across surveys. In addition, the outcome of those comparisons largely depended on the variables included in the estimation. The observed sample differences are remarkable when considering that three nonprobability surveys were drawn from the same online panel. Adjusting the nonprobability surveys somewhat improved their accuracy.
- 28 September 2024
- Published by SAGE Publications in Sociological Methodology
- Vol. 55 (1) , 121-154
Open AccessComparing the Accuracy of Univariate, Bivariate, and Multivariate Estimates across Probability and Nonprobability Surveys with Population Benchmarks
Previous studies have shown many instances where nonprobability surveys were not as accurate as probability surveys. However, because of their cost advantages, nonprobability surveys are widely used, and there is much debate over the appropriate settings for their use. To contribute to this debate, we evaluate the accuracy of nonprobability surveys by investigating the common claim that estimates of relationships are more robust to sample bias than means or proportions. We compare demographic, attitudinal, and behavioral variables across eight German probability and nonprobability surveys with demographic and political benchmarks from the microcensus and a high-quality, face-to-face survey. In the analyses, we compare three types of statistical inference: univariate estimates, bivariate Pearson’s r coefficients, and 24 different multiple regression models. The results indicate that in univariate comparisons, nonprobability surveys were clearly less accurate than probability surveys when compared with the population benchmarks. These differences in accuracy were smaller in the bivariate and the multivariate comparisons across surveys. In addition, the outcome of those comparisons largely depended on the variables included in the estimation. The observed sample differences are remarkable when considering that three nonprobability surveys were drawn from the same online panel. Adjusting the nonprobability surveys somewhat improved their accuracy.
- Open Access
- 4 March 2023
- Published by Center for Open Science in SocArXiv
Comparing the Accuracy of Univariate, Bivariate, and Multivariate Estimates across Probability and Non-Probability Surveys with Population Benchmarks
Previous studies have shown many instances where non-probability surveys were not as accurate as probability surveys. However, because of their cost advantages, non-probability surveys are widely used, and it is debated for what settings they are appropriate. To contribute to this debate, we evaluate the accuracy of non-probability surveys by investigating the common claim that estimates of relationships are more robust to sample bias than means or proportions. We compare demographic, attitudinal, and behavioral variables across eight German probability and non-probability surveys with demographic and political benchmarks from the Microcensus and a high-quality face-to-face survey. In the analyses, we compare three types of statistical inference: univariate estimates, bivariate Pearson’s r coefficients, and 24 different multiple regression models. The results indicate that in univariate comparisons, non-probability surveys were clearly less accurate than probability surveys when compared with the population benchmarks. These differences in accuracy were smaller in the bivariate and the multivariate comparisons across surveys. In addition, the outcome of those comparisons largely depended on the variables included in the estimation. The observed sample differences are remarkable when considering that three non-probability surveys were drawn from the same online panel. Adjusting the non-probability surveys somewhat improved their accuracy.
- 4 March 2023
- Published by Center for Open Science in SocArXiv
Open AccessComparing the Accuracy of Univariate, Bivariate, and Multivariate Estimates across Probability and Non-Probability Surveys with Population Benchmarks
Previous studies have shown many instances where non-probability surveys were not as accurate as probability surveys. However, because of their cost advantages, non-probability surveys are widely used, and it is debated for what settings they are appropriate. To contribute to this debate, we evaluate the accuracy of non-probability surveys by investigating the common claim that estimates of relationships are more robust to sample bias than means or proportions. We compare demographic, attitudinal, and behavioral variables across eight German probability and non-probability surveys with demographic and political benchmarks from the Microcensus and a high-quality face-to-face survey. In the analyses, we compare three types of statistical inference: univariate estimates, bivariate Pearson’s r coefficients, and 24 different multiple regression models. The results indicate that in univariate comparisons, non-probability surveys were clearly less accurate than probability surveys when compared with the population benchmarks. These differences in accuracy were smaller in the bivariate and the multivariate comparisons across surveys. In addition, the outcome of those comparisons largely depended on the variables included in the estimation. The observed sample differences are remarkable when considering that three non-probability surveys were drawn from the same online panel. Adjusting the non-probability surveys somewhat improved their accuracy.