In this article, the authors introduce a novel statistical method to measure persuasion within small groups.
Construct & Measurement: "Persuasion" is defined as an individual's systematic preference change resulting from interpersonal interaction—separate from measurement error. The approach measures two types of persuasion: one in a latent (left-right) preference space and another in topic-specific policy spaces.
Experimental Illustration: This method analyzes data from a large-scale randomized deliberative field experiment focused on US fiscal policy preferences. Participants were randomly assigned to discussion groups, allowing researchers to assess how group composition influences opinion shifts.
Key Findings & Implications: Results challenge the "law of small-group polarization," which typically shows views moving more extreme through group deliberation. Instead, findings reveal systematic persuasion effects—aligning with deliberative aspirations—in both latent and policy-specific domains.