Reading focus
Syllabus: How Can We Study Experience and Meaning?
Connect private responses, group discussion and observed decisions, including the move from translated transcripts to coded categories. Separate their interpretive roles from the randomized comparison.
Why you’re reading it. Assess what mixed evidence contributes and which proposed processes remain untested.
A question to consider. Identify one conclusion that needs the qualitative material and one causal claim it cannot establish. How would your interpretation change if public and private responses diverged?
Why this reading
Paluck tests a reconciliation radio drama in Rwandan communities and distinguishes changes in personal beliefs, perceived norms and observed behavior. The field experiment makes media influence a question about participation in a social setting, not simply exposure to a persuasive message.
Before you read
Twelve communities were paired and assigned to listen to either a reconciliation or a health radio drama for a year. Assignment occurred at the community level. People listened in organized groups, so the intervention combines a programme with a shared listening experience.
Read for
- Distinguish measures of personal belief, perceived norms and behavior, noting which outcomes changed and which did not.
- Follow the matched-community design and the evidence from interviews, group discussion and the allocation of a shared resource.
Consider afterwards
- How could an experiment distinguish the contribution of the story from the contribution of discussing it with others?
- What would count as evidence that a public change in behavior persists when people are no longer being observed?
Editorial context
The study’s small number of assigned communities, dedicated listening groups and particular post-genocide political setting constrain generalization. Many personal beliefs changed little, and effects were not uniform across all measured norms and behaviors. Group discussion and emotional involvement are plausible processes, but they were not independently randomized. The study therefore cannot establish a simple causal chain from changed norms to changed behavior, or demonstrate long-term prevention of violence. The author calls for more precise measures and tests of duration and mechanisms.