Degree

Doctor of Philosophy (PhD)

Department

Psychology

Document Type

Dissertation

Abstract

Despite decades of evidence demonstrating the predictive validity of structured interviews and cognitive ability tests, practitioners continue to rely on less effective methods such as unstructured interviews. One explanation is that practitioners struggle to understand traditional statistical indices (e.g., correlations), leading to poor cue utilization and overconfidence in intuition. This study examines whether alternative validity displays (icon arrays, binomial effect size displays, and expectancy charts) improve comprehension, reduce overconfidence, and promote evidence-based decision-making in personnel selection. Guided by the Brunswik Lens Model, dual-process theory, and dual coding theory, participants evaluated applicants across conditions varying in display format both in real time and after a time delay. Results revealed mixed evidence regarding the influence of display format on overconfidence, comprehension, calibration, and decision accuracy. Importantly, organizations may benefit from pairing accessible validity displays with training, decision aids, and structured procedures that ensure decision-makers understand and appropriately apply validity evidence. More broadly, the findings highlight comprehension as a critical mechanism for reducing reliance on intuition and improving evidence-based personnel selection.

Date

9-20-2026

Committee Chair

Don Zhang

LSU Acknowledgement

1

LSU Accessibility Acknowledgment

1

Download

Available for download on Sunday, October 02, 2033

Share

COinS