Under What Conditions Are Observed Intercorrelations Greater Or Smaller Than True Intercorrelations?

Kevin R. Murphy, Robert Jako

Research output: Contribution to journalArticlepeer-review

Abstract

Two possible explanations were examined for finding that observed correlations among dimensions are sometimes greater than and sometimes less than true intercorrelations: It can be explained by (a) variation in true halo or (b) in terms of individual differences in perceived similarities among dimensions. Subjects (N = 145) viewed videotapes that exhibited either high or low true halo, rated them, and rated similarities among dimensions. True halo level accounted for significant variance in rating intercorrelation, but perceived similarity did not. Results suggest that neither explanation is adequate to account for the finding that observed intercorrelations can either overestimate or underestimate true intercorrelations.

Original languageEnglish
Pages (from-to)827-830
Number of pages4
JournalJournal of Applied Psychology
Volume74
Issue number5
DOIs
Publication statusPublished - Oct 1989
Externally publishedYes

Fingerprint

Dive into the research topics of 'Under What Conditions Are Observed Intercorrelations Greater Or Smaller Than True Intercorrelations?'. Together they form a unique fingerprint.

Cite this