We plan to implement a supplementary online survey as a follow-up to findings from a previous study that examines the link between experimentally elicited measures of biases in social behavior towards people with different religious affiliations of religious leaders and congregants from their churches in the Busia county, Kenya. We find evidence of discrimination towards Muslims and non-religious people, and substantial heterogeneity – some people are tolerant and some do discriminate. In the planned supplementary online survey, we want to experimentally measure the same type of preferences and learn more about perceptions about people with different religious affiliations.
Intervention
Not available at this moment.
Experimental design
Not available at this moment.
Design details
Randomization method
The experimental conditions will be implemented between-subject (money allocation task) and within-subject (survey questions). The order of the decisions is randomized by Qualtrics.
In the control condition, participants will make choices affecting a Christian recipient (and answer questions measuring their perceptions of Christians), and in the two treatment (religious out-group) conditions they will make choices affecting a Muslim recipient and a non-religious recipient (and answer questions measuring their perceptions of Muslims and non-religious people).
Randomization unit
The experimental conditions will be implemented between-subject at the individual level (money allocation task) and within-subject (survey questions).
In the control condition, participants will make choices affecting a Christian recipient (and answer questions measuring their perceptions of Christians), and in the two treatment (religious out-group) conditions they will make choices affecting a Muslim recipient and a non-religious recipient (and answer questions measuring their perceptions of Muslims and non-religious people).
Clusters
500 individuals recruited by a survey agency, Kenyan adults, Christians, diverse in terms of gender, age, rural/urban residence and education.
Observations
500 individuals recruited by a survey agency, Kenyan adults, Christians, diverse in terms of gender, age, rural/urban residence and education.
Arms
Total sample size will be 500 individuals. In the money allocation task, which will be implemented between-subject, there will be approximately 167 participants in each experimental condition (Control condition - Christian recipient, Treatment 1 condition - Muslim recipient, Treatment 2 condition - non-religious recipient).
Primary outcomes
(i) Allocations in the Help-or-Harm tasks towards recipients with different religious affiliation (Christian, Muslim, non-religious). (ii) Choices in a set of questions aiming to measure perceptions of wealth, morality, threat and personal interactions.
Cited by
No citation data available for this registration yet.