SELECT – Judgment in Selection Settings (SELECT)

Basic data for this project

Description

We are interested in how decision makers form judgments in real-life selection settings. Whether hiring new employees or deciding on academic admissions, selection processes have enormous consequences for all involved parties. While there has been much research on the validity of different selection measures and variables (e.g., high predictive validity for cognitive ability, structured interviews, conscientiousness; low predictive validity for graphology, unstructured interviews, perceived attractiveness), there is little insight on how decision makers combine available information. We do know however, that an intuitive combination of different measures and variables leads to a substantial loss of validity compared to a pure mechanical combination. Thus, in the SELECT study we aim at investigating how individuals combine different sources of information,what judgment processes underlie these decisions across and within different selection procedures, and why an intuitive combination of information leads to a loss of validity. In more detail, we will answer the following research questions: To what degree do judges use valid and non-valid information? How consistent do judges use given information? Do personality and / or experience of judges explain differences in judgments? To answer these questions, we conduct online studies and a real-life assessment center study. Online studies: In two online studies, we send out real applicant profiles (Study 1A) and real applicants CVs (Study 1B) to experts and non-experts. Doing so, we include valid information (e.g., intelligence) as well as non-valid information (perceived liking) and investigate judgment processes. Real-life study: In cooperation with the medical department of the University of Münster we analyze medical school assessment center data. For this, we code performance videos of more than 100 real applicants (based on a comprehensive coding scheme) and analyze judgments by more than 40 decision makers (medical doctors). This enables us to specifically investigate what cues are used when judging performance and how consistent decision makers use those cues.