Integrating Multiple Informants’ Reports: How Conceptual and Measurement Models May Address Long-Standing Problems in Clinical Decision-Making

Abstract

Assessing youth psychopathology involves collecting multiple informants’ reports. Yet multi-informant reports often disagree, which necessitates integrative strategies that optimize predictive power. The trait-score approach leverages principal components analysis to account for the context and perspective from which informants provide reports. This approach may boost the predictive power of multi-informant reports and thus warrants rigorous testing. We tested the trait score approach using multi-informant reports of adolescent social anxiety in a mixed clinical and community sample of adolescents ( N = 127). The trait score incrementally predicted observed social anxiety (βs = 0.47–0.67) and referral status (odds ratios = 2.66–6.53) above and beyond individual informants’ reports and a composite of informants’ reports. The trait score predicted observed behavior at magnitudes well above those typically observed for individual informants’ reports of internalizing psychopathology (i.e., rs = .01–.15). Findings demonstrate the ability of the trait score to improve prediction of clinical indices and potentially transform widely used practices in multi-informant assessments.

Longitudinal Associations between Callous-unemotional (CU) Traits and School-based Affiliative Relationships among South Korean Children

“…First, we relied on child report for both CU traits and teacher-child relationship quality, which may lead to shared method variance. Further, the differing perspectives of multiple informants who interact with the child in different contexts can provide useful information (Kraemer et al, 2003;Makol et al, 2020). For example, child covert and deceitful antisocial acts, especially those associated with CU traits, may not be detected by adult informants.…”

Phenotypic and genetic markers of psychopathology in a population-based sample of older adults

“…Indeed, we observed a high correlation between the internalizing and externalizing factors in the HRS sample, which is thought to indicate the presence of a higher-order bifactor 13 , 85 . Moreover, the gold standard for measuring psychiatric symptoms and disorders is through structured clinical interviews—e.g., the Structured Clinical Interview for DSM Disorders 86 —or via questionnaires administered to multiple informants 87 . For example, several previous investigations of the structure of psychopathology in younger samples 14 , 16 have relied on structured clinical interviews to measure symptoms of MDD, generalized anxiety disorder, multiple types of phobias, and panic disorder for internalizing psychopathology, and symptoms of alcohol use disorder, drug use disorder, conduct disorder, and antisocial personality disorders for externalizing psychopathology.…”

How robust is the p-factor? Using multitrait-multimethod modeling to inform the meaning of general factors of psychopathology in youth

“…CTOM (and CTCM) models are most appropriate in this context given that evaluative consistency bias is conceptualized as unidimensional. Second, the CTOM model is conceptually most compatible with other widely used methods to model multi-informant youth psychopathology data (e.g., Bauer et al, 2013;Kraemer et al, 2003;Makol et al, 2020). Third, P-FACTOR AND MULTITRAIT-MULTIMETHOD MODELING 21 the CTCU model is among the most widely criticized MTMM models (Lance et al, 2002).…”

“…Moreover, unique variance attributable to each informant is consistently associated with theoretically relevant external criteria, including above and beyond what is shared across informants as well as unique to other informants. Previous work has shown that youth, parent, and confederate reports of adolescent social anxiety demonstrate incremental validity above and beyond one another in the statistical prediction of behavioral indicators during social interaction tasks, whether the adolescent received mental health services, clinical severity, and the like (Cannon et al, 2020;Glenn et al, 2019;Makol et al, 2020Makol et al, , 2021. Informant divergence is also associated with poorer outcomes for certain domains.…”