View article

This study revisits the relationship between interviews and cognitive ability tests, finding lower magnitudes of correlation than have previous meta‐analyses; a finding that has implications for both the construct and incremental validity of the interview. Our lower estimates of this relationship than previous meta‐analyses were mainly due to (a) an updated set of studies, (b) exclusion of samples in which interviewers potentially had access to applicants' cognitive test scores, and (c) attention to specific range restriction mechanisms that allowed us to identify a sizable subset of studies for which range restriction could be accurately accounted. Moderator analysis results were similar to previous meta‐analyses, but magnitudes of correlation were generally lower than in previous meta‐analyses. Findings have implications for the construct and incremental validity of interviews, and meta‐analytic methodology in general.