DEVELOPING AND VALIDATING A SECOND-ORDER REFLECTIVE FRAMEWORK FOR SCIENTIFIC CREATIVITY COMPETENCE IN HIGH SCHOOL STUDENTS: A DELPHI-SEM MIXED-METHODS APPROACH

Abstract

Background and Purpose: Scientific creativity competence (SCC) is recognized as a core higher-order competence in contemporary science education, yet validated frameworks that specify observable behavioral indicators within culturally situated, physics-based STEM learning contexts remain underdeveloped. The present study aimed to develop and validate a second-order reflective SCC measurement framework for Vietnamese upper-secondary students. Design and Methods: A sequential mixed-methods design was employed across five procedural steps: (1) systematic bibliometric synthesis of 690 Scopus-indexed publications (1960-2026); (2) second-order reflective model specification; (3) two-round Delphi expert consultation (Round 1: N = 42; Round 2: N = 38, retention rate 90.5%); (4) PLS-SEM exploratory validation (n = 300, SmartPLS 4, 5,000 bootstrap resamples); and (5) CB-SEM confirmatory validation incorporating EFA and CFA (n = 1,200, SPSS 27 and AMOS 27). Quantitative participants were Grade 10 students recruited from schools in Ho Chi Minh City and Gia Lai Province via stratified sampling. Results: The proposed framework - comprising four first-order components (PRB: problem recognition; IDE: idea generation; DES: solution design; EVA: solution evaluation) operationalized through 16 behavioral indicators - achieved expert consensus rates of 86.8%-92.1% in Round 2. PLS-SEM confirmed adequate internal consistency (Cronbach’s α = 0.752-0.792; CR = 0.843-0.865), convergent validity (AVE = 0.575-0.617), and discriminant validity (HTMT = 0.605-0.771). CFA demonstrated excellent model fit - χ²(100) = 97.892, p = 0.541; χ²/df = 0.979; GFI = 0.990; CFI = 1.000; TLI = 1.000; RMSEA = 0.000; SRMR = 0.017; PCLOSE = 1.000 - and confirmed the second-order reflective structure, with DES and PRB as the dominant reflections of SCC (β = 0.85; R² = 0.72), followed by IDE (β = 0.74) and EVA (β = 0.71). All three research hypotheses were supported. Conclusions and Implications: The validated SCC framework constitutes, to the best of the authors’ knowledge, the first psychometrically rigorous second-order measurement model for scientific creativity competence developed and normed within the Vietnamese secondary education context. The five-step Delphi-SEM procedure offers a replicable protocol for competence framework development in educational measurement. The framework provides a theoretically grounded foundation for designing assessment instruments and SCC-targeted STEM learning sequences in Physics instruction at the upper-secondary level.