Quantifying talk: developing reliable measures of verbal productivity

Behav Res Methods. 2011 Mar;43(1):168-78. doi: 10.3758/s13428-010-0019-y.

Abstract

Measuring talkativeness is of interest to several areas of research. However, there are few brief, validated measures available. We examined test-retest reliability, inter-relationships and convergent/divergent validity for five brief measures of verbal productivity. Nineteen men and 32 women participated in four sessions, completing five speech tasks that varied in demand, purpose of speech and sociability. Several potential metrics (word count, duration and rate) were examined. All tasks except a novel Unprompted Speech task demonstrated good word count test-retest reliability (interclass correlation coefficients from .71 to .85). Factor analysis revealed low-demand, non-functional tasks formed one factor ("Voluntary Talkativeness"), while higher demand tasks formed a second factor ("Speech Ability"). This finding and examination of relationships with IQ, personality and gender indicate "Voluntary Talkativeness" is not wholly accounted for by verbal ability, and is only weakly related to self-reported personality. Recommendations for the measurement of "Voluntary Talkativeness" are made.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Adolescent
  • Adult
  • Communication
  • Ethnic Groups
  • Factor Analysis, Statistical
  • Female
  • Humans
  • Individuality
  • Intelligence
  • Male
  • Personality
  • Psychomotor Performance
  • Reproducibility of Results
  • Sex Characteristics
  • Social Behavior*
  • Speech / physiology*
  • Word Association Tests
  • Writing
  • Young Adult