Study sampling may proxy for the heterogeneity that drives deliberation outcomes.

Evidence

Secondly, we coded the sampling technique, and we distinguish between those studies which use (quasi-) random population samples (possibly quota corrected), targeted samples (focusing on specific groups of students, or people in certain geographic locations), and pure self-selection (where people volunteered to participate without stringent inclusion criteria). … These findings can be explained in two ways. On the one hand, it might be that (quasi-) random samples encompass a more diverse subset of the population. Having such a diverse subset deliberate, might have the same effect as a deliberately heterogeneously composed group. Targeted samples might be more homogeneous and might therefore lead more often to an increase in polarization. In other words, the sampling technique might be a proxy for group heterogeneity. On the other hand, the findings could be an artifact of the discipline in which the studies were conducted: in line with what we reported before, political science studies (most often using some kind of random population sample in an effort to generate external validity) primarily find that deliberation reduced polarization, whereas studies in psychology (more often using targeted samples to maximize experimental control) find that the process increases polarization. — Didier Caluwaerts et al., Deliberation and Polarization, 118

Categorization

• XREF: Connects directly to the reader’s inquiry about whether homogeneous belonging (vs. heterogeneous independence) conditions whether deliberation depolarizes or polarizes - the sampling-as-heterogeneity-proxy point speaks to what makes a genuine deliberative-democracy framing. — connects to: Studying tiny, dispersed populations forces researchers into biased sampling strategies. — Naheed Ahmed, Measurement of Perceived Interpersonal…, 112

Broader Connections

• Studying tiny, dispersed populations forces researchers into biased sampling strategies. #expands (0.78) — Naheed Ahmed, Measurement of Perceived Interpersonal…, 112