<![CDATA[Society 5.0 demands individuals who demonstrate integrated character, knowledge, and skills to navigate technology-driven environments. However, existing research rarely compares different estimation strategies in Confirmatory Factor Analysis (CFA) when modeling competency structures. This study aims to identify stable competency indicators for the Society 5.0 era (C-5.0) using several CFA estimation approaches across ordinal and continuous data types. A quantitative survey was conducted with Indonesian TVET academics and practitioners, and data were analyzed using multiple statistical simulations. The model included 30 manifest indicators representing affective, cognitive, and psychomotor domains. Findings reveal that only six indicators consistently remained robust across estimation techniques: curiosity, environmental care, and social care (character domain); creative thinking and innovative thinking (knowledge domain); and cooperative skills (skills domain). One indicator—critical thinking—did not satisfy goodness-of-fit criteria in any estimation scenario, suggesting that its contribution may be context-dependent within TVET environments. These findings highlight that not all theoretically important indicators operate similarly under different measurement conditions. Methodologically, the study emphasizes the importance of testing alternative CFA estimation strategies when competency frameworks involve mixed data types. Practically, the results provide policymakers and vocational institutions with empirically validated priorities for curriculum development, focusing on competencies that demonstrate measurement stability in real-world contexts. Future studies should examine longitudinal validation and cross-cultural replication to strengthen generalizability.]]>
