Interviewer accuracy across levels of structure in the employment interview

Abstract

It is now common knowledge that more structured interviews result in greater criterion validity with respect to predicting performance. Theorists and researchers have also noted that interviewers differ in their ability to judge others. This study examined the relationships between interview structure, observer intelligence, and observer concept of the ideal candidate and observer accuracy. This study proposed that: 1) Greater interview structure would lead to greater observer accuracy; 2) General intelligence would moderate the relationship between interview structure and accuracy; and 3) Observers' concept of the ideal candidate would moderate the relationship between interview structure and accuracy. Three hundred and one undergraduate psychology students viewed mock employment interviews conducted with attitudinal questions, situational questions, and past-behavioral questions and rated candidate responses, using a behaviorally-anchored rating scale. Only the first hypothesis received limited support. Accuracy was significantly different among three interview conditions: an attitudinal interview (AI), situational interview (SI), and patterned behavioral description interview (PBDI). Accuracy was greatest in the SI, followed by the AI, and least in the PBDI. A small significant positive effect was found between intelligence and accuracy, and a small negative effect was found between dissimilarity of observers' prototype from an ideal candidate (as established by the O*NET) and accuracy, after controlling for interview condition, but no support was found for the proposed interaction effects. Possible reasons for theses results are explored and discussed.