Sources of systematic error in panel job interview: A Case Study of a Public University

The objective of this study is to investigate the sources of systematic errors associated with the scores obtained from interviews of job applicants and to estimate the generalizability and dependability coefficients within the framework of generalizability theory. Accordingly, the data on the job competency assessment scores of the applicants for the three positions of one the public universities were analyzed using a two-way, fully crossed design. Then, the regular sources of error, including the questions, interviewers and interaction of errors were determined and measured, with the help of which the generalizability and dependability coefficients of the scores were estimated. The findings suggest that generalizability and dependability coefficients of the assessment scores are not equal in all job interviews. Moreover, since the generalizability and dependability coefficients are less than expected, the mean assessment score of one job should not be the basis for decision-making for all the three job selections. Based on the results, it is suggested that, before the adoption of the interviewer's opinions about the merits of applicants, the generalizability and dependability coefficients of the scores be estimated using generalizability theory and if the coefficients are low, the job interview scores should not be the basis for decision-making.