Situational judgment tests in high-stakes settings: issues and strategies with generating alternate forms

This study used principles underlying item generation theory to posit competing perspectives about which features of situational judgment tests might enhance or impede consistent measurement across repeat test administrations. This led to 3 alternate-form development approaches (random assignment, incident isomorphism, and item isomorphism). The effects of these approaches on alternate-form consistency, mean score changes, and criterion-related validity were examined in a high-stakes context (N = 3,361). Generally, results revealed that even small changes in the context of the situations presented resulted in significantly lower alternate-form consistency. Conversely, placing more constraints on the alternate-form development process proved beneficial. The contributions, implications, and limitations of these results for the development of situational judgment tests and high-stakes testing are discussed.