Inequality in talk and group size effects: An analysis of measures

Mary R. Rose*, Shari Seidman Diamond, Daniel A. Powers

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


The earliest studies of talk in small groups indicated that larger groups experience more inequality in participation than smaller groups. However, there has been insufficient attention to how to properly measure inequality when group size varies. We describe properties of a common inequality metric, the Gini coefficient, and consider it in light of early efforts that modeled talk in small groups using harmonic and exponential distributions. We use these classic distributions to develop novel inequality measures and also consider a measure developed specifically to examine inequality across small systems of different sizes (the CON). We apply all measures of inequality to data from four highly realistic jury deliberation datasets, including one involving real juries, examining both word counts and turns. All indicators correlate very highly with one another, but both the Gini and a Gini adjusted for group size privilege smaller groups over larger ones, producing significant positive correlations with group size. The model-based values and the CON offer a different ordering of datasets compared to the Gini and do not show the same correlations with group size. Results offer several reasons to recommend the CON measure as a promising way of comparing inequality across small groups of different sizes.

Original languageEnglish (US)
Pages (from-to)778-798
Number of pages21
JournalGroup Processes and Intergroup Relations
Issue number5
StatePublished - Aug 1 2020


  • group size
  • inequality
  • juries
  • measurement
  • participation
  • small groups

ASJC Scopus subject areas

  • Social Psychology
  • Cultural Studies
  • Communication
  • Arts and Humanities (miscellaneous)
  • Sociology and Political Science


Dive into the research topics of 'Inequality in talk and group size effects: An analysis of measures'. Together they form a unique fingerprint.

Cite this