The extent to which ratings of interviewee performance reflect certain constructs varies widely depending on the level of structure of the interview, the kind of questions asked, interviewer or applicant biases, applicant professional dress or nonverbal behavior, and a host of other factors. For example, some research suggests that applicant’s cognitive ability, education, training, and work experiences may be better captured in unstructured interviews, whereas applicant’s job knowledge, organizational fit, interpersonal skills, and applied knowledge may be better captured in a structured interview.
Further, interviews are typically designed to assess a number of constructs. Given the social nature of the interview, applicant responses to interview questions and interviewer evaluations of those responses are sometimes influenced by constructs beyond those the questions were intended to assess, making it extremely difficult to tease out the specific constructs measured during the interview.Reducing the number of constructs the interview is intended to assess may help mitigate this issue. Moreover, of practical importance is whether the interview is a better measure of some constructs in comparison to paper and pencil tests of the same constructs. Indeed, certain constructs (mental ability and skills, experience) may be better measured with paper and pencil tests than during the interview, whereas personality-related constructs seem to be better measured during the interview in comparison to paper and pencil tests of the same personality constructs. In sum, the following is recommended: Interviews should be developed to assess the job relevant constructs identified in the job analysis
