Candidate results from multiple employee selection procedures are most often combined clinically using hiring managers’ judgment, but evidence suggests this approach attenuates predictive validity compared to mechanical data combination approaches due to hiring managers’ misconceptions about selection procedures. The present research proposes and tests a model that explains how and why hiring managers determine the extent to which they will utilize candidate results from various selection procedures or sources of candidate information. Specifically, the model posits that utilization is driven by user beliefs about the predictiveness and fairness of procedures, which are in turn informed by nine dimensions of “user reactions” or perceptions of properties of predictor methods (procedural autonomy, evaluation autonomy, fidelity, fakability, evaluation consistency, and transparency) and predictor constructs (job relatedness, malleability, and development equity). Study 1 tested the model by asking a sample of hiring managers to view and rate selection procedures manipulated into nine predictor method/construct combinations. Results indicated that users’ predictiveness beliefs are a stronger driver of utilization than fairness beliefs; additionally, most dimensions of user reactions predicted utilization intentions in the hypothesized direction. Study 2 evaluated the efficacy of autonomy-based interventions by having a sample of hiring managers view and rate a structured interview or computerized assessment manipulated into high and low levels of autonomy. Results suggested that slightly increasing hiring manager autonomy within standardized selection procedures led to increases in predictiveness beliefs and utilization intentions. Findings have implications for designing selection procedures and tailoring hiring manager communication and training efforts, both with the goal of encouraging utilization of more valid predictors in employee selection decision-making contexts.
Read
