Situational Interviews and Job Performance: The Results in One Public Agency

Article excerpt

This article describes a recent effort by the Mississippi Department of Corrections to higher quality employees through a situational interview format. A comparison of correctional officers chosen through the situational interview to those chosen through traditional selection procedures reveals little difference between the two groups. The situational interview is not associated with systematic improvements in job performance scores, absenteeism rates, or attitudes toward management as might be expected. Hence the value of this technique in terms of its predictive validity is brought into question.

Gerald T. Gabris is an associate professor of Public Adminstraat Northern Illinois University. He has published numerous articles in the areas of origanizational change and performance, and services on the editorial board of two public administration journals.

This article investigates whether employees selected on the basis of a "situational" interview perform better and exhibit more positive attitudes toward management compared to those hired through more conventional unstructured interview selection procedures. If situational interviews systematically identify better employees, and this identification of high performance does not decay over time, then the technique should be recommended for broader use in the public sector.

An unique aspect of this study is the evaluation of employee performance over time based on different interview selection techniques. Conceivably, a selection procedure may identify preferable behavior attributes among employee candidates during initial screening, only to find these differentials gradually overwhelmed by forces connected with organizational culture and socialization. Whether employee selection methods predict durable employee performance pattern is an important question, since organizations expend considerable resources constructing professionally developed selection procedures. Do these formal and elegant mechanisms really make a difference over the long haul?

To explore these possibilities, a sample of employees from the Mississippi Department of Corrections (MDOC) main prison facility win be examined. The MDOC currently uses the situational interview as the final selection hurdle for entry level correctional officers. The situational interview signifies an attempt by the Agency to increase the quality and performance of correctional personnel. Presumably, the process will identify career-oriented individuals with higher potentials for professionalism and job performance. The question is whether this type of selection tool facilitates these management objectives.

Issues Surrounding Selection Procedures

The purpose of employee selection is to increase the probability of choosing successful job candidates (Guion, 1966; Arvey and Faley, 1988). If a test or interview does not predict job success at a rate greater than chance, then the business utility of the instrument is questionable.

Selection procedures also automatically involve a complex panoply of legal issues, many emanating from the case of Griggs v. Duke Power Company (1971). This precedent-setting decision established the notion that the "effect" of a hiring procedure may result in unfair discrimination (i.e., that persons not selected on the basis of some test score could have equivalent opportunities for job success as those selected by the test). Griggs also stipulated that tests must be job related and not abstract, and that if the plaintiff demonstrates disparate impact, the burden of persuasion for indicating how a selection procedure is valid shifts to the defendant. Moreover, even if a defendant can validate a test, it must adopt less discriminatory selection procedures if such procedures exist.(1) Employee selection is likely to remain a high stakes game for most employers where slight mistakes could lead to several and costly penalties.(2) The discussion suggests that the utilization of professionally developed tests relies on two key assumptions. …