Recruiters prefer expert recommendations over digital hiring algorithm: a choice-based conjoint study in a pre-employment screening scenario

Abstract

Purpose This study aims to analyze aspects of decision-making in recruitment. Using a choice-based conjoint (CBC) experiment with typified screening scenarios, it was analyzed what aspects will be more important for recruiters: the recommendation provided by a hiring algorithm or the recommendation of a human co-worker; gender of the candidate and of the recruiter was taken into account. Design/methodology/approach A total of 135 recruitment professionals (67 female) completed a measure of sex roles and a set of 20 CBC trials on the hiring of a pharmacologist. Findings Participants were willing to accept a lower algorithm score if the level of the human recommendation was maximum, indicating a preference for the co-worker’s recommendation over that of the hiring algorithm. The biological sex of neither the candidate nor the participant influenced in the decision. Research limitations/implications Participants were presented with a fictitious scenario that did not involve real choices with real consequences. In a real-life setting, considerably more variables influence hiring decisions. Practical implications Results show that there are limits on the acceptance of technology based on artificial intelligence in the field of recruitment, which has relevance more broadly for the psychological correlates of the acceptance of the technology. Originality/value An additional value is the use of a methodological approach (CBC) with high ecological validity that may be useful in other psychological studies of decision-making in management.