A tale of two formats: Direct comparison of matching situational and behavior description interview questions

The primary purpose of this meta-analysis was to examine the psychometric properties of situational interview (SI) and behavior description interview (BDI) questions written to assess the same set of job attributes. Using a final dataset of 29 coefficients (N = 8148), we found an observed mean correlation of 0.40 (0.47 corrected) between construct-matched SI and BDI questions. In terms of moderators, even lower correspondence resulted when (1) both question types had lower internal consistency, (2) there were more questions per format, (3) probing was not allowed in either format or when allowed in one format but not in the other, and (4) the purpose of the interview was for research (vs. employment). Given that their correspondence can be quite low under some moderator combinations, an important implication is that SI and BDI questions should not automatically be assumed to be interchangeable, even when written deliberately to assess the same attributes, and that incremental validity is very possible. Further, results suggest that SIs have higher overall mean validity compared to BDIs for predicting job performance (0.23 vs. 0.18, respectively). Results also indicated a slightly stronger relationship with cognitive ability measures for BDIs than SIs (0.11 vs. 0.09, respectively).