Cross-rater agreement on common and specific variance of personality scales and items

René Mõttus*, Robert R. McCrae, Jüri Allik, Anu Realo

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Using the NEO Personality Inventory-3, we analyzed self/informant agreement on personality traits at three levels that were made statistically independent from each other: domains, facets, and individual items. Cross-rater correlations for the common variance in the five domains ranged from 0.36 to 0.65 (M = 0.49), whereas estimates for the specific variance of the 30 facets ranged from 0.40 to 0.73 (M = 0.56). Cross-rater correlations of residual variance of individual items ranged from -0.14 to 0.49 (M = 0.15; 88% statistically significant at p< 0.002). Agreement on common variance was moderately related to item observability and evaluativeness, whereas variance played a larger role. Facets and even single items detect nuances of personality variation that may merit substantive attention.

Original languageEnglish
Pages (from-to)47-54
Number of pages8
JournalJournal of Research in Personality
Volume52
DOIs
Publication statusPublished - 1 Oct 2014

Keywords / Materials (for Non-textual outputs)

  • Bi-factor
  • Consensual validity
  • Corss-rater agreement
  • Cross-informant agreement
  • Cross-observer agreement
  • Nuances
  • Specific variance

Fingerprint

Dive into the research topics of 'Cross-rater agreement on common and specific variance of personality scales and items'. Together they form a unique fingerprint.

Cite this