Using Generalizability Theory to Estimate the Dependability of Individual Cut Scores within Biomedical and Other Assessments

In this brief article, we describe how generalizability theory can be used to derive indices representing the dependability of cut scores used in decision making based on data from one or two occasions. We include an empirical example using the Neuroticism subscale from the Big Five Inventory to illustrate use of these coefficients and how the dependability of cut scores can exceed the overall dependability of scores when considered on whole. We also direct readers to resources with computer code and further information about conducting the illustrated analyses.