Evaluating Summary Measures of Heterogeneous Political Discussion: The Critical Roles of Excluded Cases and Discussion with People Holding Extreme Views

Abstract

Summary network heterogeneity measures are frequently used and often-cited in work on interpersonal political disagreement, but their properties are not well documented and they produce anomalous results relative to other measures of socially supplied disagreement. This study deconstructs the familiar summary network heterogeneity measure to examine why it produces incongruous results relative to other measures used in the literature. This study pays particular attention to the multiple strategies that exist for handling non-partisans when creating the measure. Additionally, it focuses on the “extreme views” items included in its construction, which are conceptually and methodologically distinct from other items and rarely theoretically justified. Results demonstrate that summary network heterogeneity is best thought of an as index, rather than as a scale, which has implications for how scholars should approach it. Critically, inclusion of the extreme views items and decisions about data exclusion also have substantive consequences; this highlights the need for further inquiry into extremity and for explicit reporting about data-handling decisions.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.