The contrast effect in a competency based situational interview

Abstract

The recruitment interview is used ubiquitously by organisations in the UK as part of the recruitment and selection process. Despite improvements over the years, the method is still prone to error and it is important for organisations to take what steps they can to reduce error in selection decisions. One source of error identified and accepted as a cause of bias within the interview is that of contrast effect. This effect causes assessors to base their judgements of candidates partly on a comparison to earlier performances of other candidates. This has the effect of giving inflated scores to interviewees when others are poor and lower scores when others are good. The presence of this effect is assumed due to various studies carried out within the experimental paradigm without any quantitative evidence collected from real world settings. This study collected data on 694 interviews carried out to recruit cabin crew for a major UK based airline. The data set provided 230 interview pairs for analysis. Correlational analyses showed that the prior performance of one candidate could significantly affect the selection outcome decision of a subsequent candidate. Binary Logistic Regression revealed the scores given to the subsequent candidate mediated this relationship. The implications for practice are discussed. These include improved rater training, the implications of interview timetabling, and rotation of assessor teams.