Key points are not available for this paper at this time.
Contemporary psychometric research is increasingly characterized by parallel developments in latent variable modelling and network psychometrics. Despite their shared goals, these approaches are rarely integrated within a single construct validation framework, limiting opportunities for methodological triangulation and robust inference. In the present study, we demonstrate a complementary psychometric workflow based on exploratory factor analysis (EFA), exploratory graph analysis (EGA), confirmatory factor analysis (CFA), construct validation, and measurement invariance testing to evaluate construct dimensionality and applicability. This approach is illustrated using the Metacognitions Questionnaire–30 (MCQ-30) as a substantive case study. Using data from 3,999 participants across student and general population samples, we examined convergence between factor-analytic and network-based dimensionality estimates, assessed item stability and structural consistency, and evaluated construct, incremental, and measurement invariance across demographic and symptom severity groups. Results showed strong convergence between EFA and EGA in identifying a correlated five-dimensional structure, which was subsequently supported in CFA. Measurement invariance was established across age, gender, and anxiety and depression severity groups, enabling meaningful latent mean comparisons. Rather than offering a scale-specific validation, this study demonstrates how integrating latent variable and network psychometric approaches can strengthen construct validation, enhance transparency, and improve the robustness of measurement conclusions. The proposed framework provides a practical illustration for psychological researchers seeking to evaluate dimensionality and measurement equivalence across diverse contexts, highlighting the complementarity between common factor and network psychometric modelling.
Anyan et al. (Mon,) studied this question.