The purpose of this dissertation research was to use an information processing approach to begin to understand how emotions influence the interview information search process. Based on information processing theory, it was hypothesized that interviewers induced into positive affect would ask fewer questions and seek positive information, while participants induced into negative affect would ask more questions and seek negative information. Secondary analyses were put forth to examine sensitivity to emotions as well as indicators of differential information processing. The hypotheses and secondary analyses were tested via a 3-group between-subjects design, where participants received either a positive affect, neutral affect, or negative affect mood induction and then performed a computer simulation of an unstructured job interview.
The primary analysis suggested that participants in the negative affect condition gave lower applicant ratings and reported a lower likelihood of recommending the applicant for hire than participants in the positive affect condition. Supporting the hypotheses, as state-based positive affect increased, participants were more likely to ask 2 of the 3 paired questions seeking positive information than the questions seeking negative information. Opposite of the hypothesized results, as state-based positive affect increased, participants were more likely to have asked every question provided for the interview. Secondary analyses showed sensitivity to emotions was a significant
moderator of some information search behaviors. However, the secondary analyses did not indicate any differential information processing among any of the three manipulation conditions.
Overall, the results did not support the information processing approach that suggested emotions affect applicant ratings via information search behaviors. Two alternative mechanisms for how emotions influence applicant ratings, mood-as-information and mood-congruent memory recall were potentially still consistent with the results. Data were unavailable to test the mood-congruent memory recall mechanism, but data were available to test and support the mood-as-information mechanism.