Psychometrics and the Neuroscience of Individual Differences: Internal Consistency Limits Between-Subjects Effects
In the clinical neuroscience literature, between-subjects differences in neural activity are presumed to reflect reliable measures—even though the psychometric properties of neural measures are almost never reported. The current article focuses on the critical importance of assessing and reporting internal consistency reliability—the homogeneity of “items” that comprise a neural “score.” We demonstrate how variability in the internal consistency of neural measures limits between-subjects (i.e., individual differences) effects. To this end, we utilize error-related brain activity (i.e., the error-related negativity or ERN) in both healthy and generalized anxiety disorder (GAD) participants to demonstrate options for psychometric analyses of neural measures; we examine between-groups differences in internal consistency, between-groups effect sizes, and between-groups discriminability (i.e., ROC analyses)—all as a function of increasing items (i.e., number of trials). Overall, internal consistency should be used to inform experimental design and the choice of neural measures in individual differences research. The internal consistency of neural measures is necessary for interpreting results and guiding progress in clinical neuroscience—and should be routinely reported in all individual differences studies.