Challenges and Ways Forward in Bridging Units of Analysis in Clinical Psychological Science

• Published in: Journal of psychopathology and clinical science 2023-10, Vol.132 (7), p.888-896
• Main Authors: Joyner, Keanan J., Perkins, Emily R.
• Format: Article
• Language: English
• Subjects: Behavioral Sciences; Experimentation; Human; Measurement; Mental disorders; Neuroimaging; Psychological Assessment; Psychopathology; Reification; Self report
• ISSN: 2769-7541; 2769-755X
• DOI: 10.1037/abn0000879

For decades, experiential measures (i.e., self- and informant-report) have dominated clinical psychological science as the primary methods of investigating the nature, causes, and consequences of psychopathology. Recent efforts to understand psychopathology in a comprehensive manner, bridging across measurement modalities and study designs, have yielded disappointing results and small cross-domain effect sizes. This problem has stymied progress in diagnosing and treating mental illness. We contend that we continue to struggle to link different units of analysis in psychopathology research because, as a field, we have not learned the lessons that have been detailed in the literature since at least the 1950s. We argue that our reliance on single units of analysis (e.g., only self-report or only neuroimaging) to conceptualize and operationalize psychological constructs lies at the heart of our struggles. Conflation of a construct with its operationalization in a single measurement modality fundamentally limits the ability to understand its manifestation in other units of analysis. To make progress toward a fine-grained understanding of the symptoms, signs, and mechanisms of psychopathology, we must move past the siloed reification of constructs and prioritize new, integrated research paradigms. General Scientific Summary This commentary highlights persistent failures of psychological science to find robust associations among data of different types, such as self-report questionnaires, behavioral task performance, or neural and physiological measures. It identifies and critiques common approaches that may contribute to these struggles to relate data of different kinds, and suggests ways forward that may contribute to more success in future investigations of this kind.