Psychometric validation of research domain criteria, general health, and clinical-psychosocial parts of the minimum dataset of the German Center for Mental Health-informing a Delphi process.

OBJECTIVE: This study evaluated the latent structure and conceptual validity of the Research Domain Criteria (RDoC), General Health, and Clinical/Psychosocial clusters of the German Center for Mental Health (DZPG) Minimum Data Set (MDS). It aimed to determine whether these clusters represent interpretable constructs and demonstrate cross-group comparability across clinical and non-clinical populations, providing a data-driven basis for item selection within a Delphi consensus process. METHODS: Two subsamples were analyzed: a representative sample (N = 2,287) and a clinic-recruited sample (N = 322) from six DZPG sites. Multi-group confirmatory factor analyses tested the latent structure and assessed configural, metric, and scalar measurement invariance between non-clinical vs. clinical/help-seeking groups. RESULTS: All three MDS clusters demonstrated excellent baseline model fit (RDoC: CFI = .990, RMSEA = .037; General Health: CFI = .993, RMSEA = .037; Clinical/Psychosocial factors: CFI = .992, RMSEA = .026). Analyses confirmed configural and scalar invariance across all clusters but indicated non-invariance at the metric level for the General Health cluster. CONCLUSIONS: The results provide initial evidence for the structural validity and cross-group comparability of the RDoC and Clinical/Psychosocial clusters, while indicating limitations regarding metric non-invariance within the General Health cluster. They informed a Delphi consensus process and support the MDS as a scalable framework for diverse populations, particularly relevant for the DZPG's effort to harmonize mental health data across institutions and contexts.