Correlational tests of predictions from a process model of the interview.

We conducted a field study to test eight propositions derived from a process model of the selection interview (Dipboye, 1982; Dipboye & Macan, 1988). According to the model, interviewers' preinterview impressions of an applicant bias the subsequent conduct of the interview and processing of information in the direction of confirming these initial impressions. To test predictions from the model, we surveyed managers and the applicants they interviewed in each of 164 interviews. In support of the model, interviewers' preinterview evaluations were positively related to postinterview evaluations of applicant qualifications and process variables predicted to mediate this relation. Results also supported the model in that interviewers with favorable preinterview impressions were more likely to attribute good interview performances to the applicants' qualifications for the job and poor performances to external factors. Contrary to the model, confidence failed to moderate the above findings, and preinterview impressions were not predictive of applicant reports of interviewers' time spent in questioning. Some possible implications of the model for future research and for improving interview practice are discussed.