Comparing The Abilities Of Structured Interviews In Predicting The Performance In Fulfilling The Task Of High Complexity

Structured interviews which consist of the situational interview (SI) and the patterned behavioral description interview (PBDI,BI), have been verified effective in predicting performance by numerous empirical researches and practice with sound validity and reliability. However, we can’t neglect two problems confronted by structured interviews. The first one is that their construct validities are still not clear, and the second one is that controversial conclusions are made as comparing the difference of situational interview and behavioral interview in predicting high-level positions. This research aims at specifying the reasons why two kinds of structured interviews are different in predicting the performance of high-level positions or tasks, and exploring what kinds of constructs are measures by structured interviews.We targeted the cooperation and leading behavior as the task of high complexity exhibited in the juniors’course of Human Resource Management. All35students enrolled in this course are selected as participants. Three studies were conducted. In the first study, we developed equivalent situational questions and behaviors questions based on same dimensions extracted from critical incident interview. In addition, a pilot study was also conducted to polish the questions and scale. In the second study, we compared the difference of structured interviews in predicting general performance, contextual performance and task performance of tasks of high complexity. What we found is that situational interview and behavioral interview are all effective in predicting general performance, and each of them respectively explains19.4%and39.3%variance of general performance. In additions, behavioral interview has incremental validity above situational interview in predicting general performance. We made further examinations to specify the reasons explaining the differences of abilities of structured interviews by establishing the relationships between structured interview and task performance, contextual performance. The results we get are that behavior interview could predict task performance and contextual performance better than situational interview, and BI has incremental validity above SI in predicting task performance, and only BI could predict contextual performance. In the third study, we attempt to examine what constructs were measured by structured interviews. By conducting correlation and regression analysis, we found that both SI and BI measured applicants’ social skills, but not personalities and self-core evaluation. SI could explain12.4%variance of social skills, and BI could explain17.1%variance of social skills.The theoretical implications of this research exist that new predicting validity of structured interview was extended and explained the why structured interviews are different in predicting tasks of high complexity in further, and made new attempts in verifying the constructs measured by structured interviews. Its practical implications embody in that theoretical explanation could be employed when employers choose appropriate interview formats to select personnel. Future study should continue comparing the difference of structured interviews in predicting high-level positions by selecting a huge amount of managers.