Toward integrative, multisource, and multimodal approaches to assessing psychopathology intergenerationally and...

Models of psychopathology contain constructs that cannot be directly observed. Assessing these constructs involves identifying specific behaviors that justifiably serve as "markers" of their presence, thus facilitating their precise and accurate measurement. Examples of these behavioral markers include the destruction of property for conduct disorder, the avoidance of unfamiliar people for social anxiety, and sleeping too much or too little for depression. These and other behavioral markers typify the measurement of both specific mental disorders as well as "p" as a transdiagnostic index of psychopathology. Importantly, researchers commonly implement measurement, methodological, and data-analytic practices that conflict with the data conditions that underlie psychopathology studies. For example, researchers often rely on one source to measure behavioral markers of psychopathology (e.g., self-report), even though 60 years of research indicate that the measurement source robustly dictates the conclusions of psychopathology studies. When researchers measure markers of psychopathology with multiple sources, they often integrate these multisource data using data-analytic strategies that misclassify valid data as measurement error; these misclassifications beget underpowered studies and, by logical extension, replication failures. These issues pertain to psychopathology research more broadly, but they also facilitate interpreting Caspi et al. (see record 2026-80066-001), and the foundation it sets for studying psychopathology both developmentally and intergenerationally. In this commentary, I highlight measurement, methodological, and data-analytic issues inspired by Caspi et al. and summarize work that supports their careful consideration. Caspi et al.'s findings call for developing integrative, multisource, and multimodal assessments that optimize the validity of psychopathology assessments intergenerationally and across the lifespan. (PsycInfo Database Record (c) 2026 APA, all rights reserved).