Low-Fidelity Simulations

Abstract

Low-fidelity simulations, which combine closed-ended response options with realistic depictions of key job tasks, have grown in popularity for several practical reasons. Research into low-fidelity simulations, which has focused primarily on text-based situational judgment tests (SJTs), has shown that such measures (a) can predict a wide range of criteria, doing so with moderate adverse impact, depending on the construct saturation of the items, (b) are no more susceptible to coaching/cheating/retesting effects than many other forms of assessment and far less susceptible than some, and (c) continue to struggle when used to measure a single construct. Practice, by contrast, has moved well beyond text-based SJTs to include multimedia SJTs, online assessment center exercises, and game-like assessments. Because these are perceived more favorably by applicants, the same research dedicated to text-based SJTs is needed to guide development of these newer forms of low-fidelity simulation.