External Psychometric Evidence for the Questionnaire of Gamification Skills in Teachers: Structural Performance, Item Limitations, and Score Interpretation in Spain

Teachers’ capacity to design gamified learning requires measures whose scores remain interpretable beyond their development context. This study examined the 22-item Questionnaire of Gamification Skills in Teachers (QGST), with wording changes from the Mexican original, in 1367 teachers in Spain. The most frequent stages were compulsory secondary (39.1%), primary (27.0%), and upper-secondary education (20.9%); categories were non-exclusive. Ordinal confirmatory factor analysis compared correlated three-factor and single-factor models, with cross-checks in two random halves. Item distributions, reliability, local residuals, and associations with specific training and implementation were examined. Separate sensitivity analyses excluded probable duplicate and constant-response records. The three-factor model outperformed the single-factor model and showed adequate comparative fit (CFI = 0.959; TLI = 0.954), but mixed absolute fit (RMSEA = 0.105; SRMR = 0.082). Four residual correlations exceeded 0.20, and omega ranged from 0.833 to 0.908. Teachers reporting training and previous implementation scored higher across dimensions, although these self-reported associations do not independently establish teaching competence. The findings support cautious research use and formative reflection through three provisional dimension scores, rather than diagnostic classification, proficiency cut-offs, or an interpretable unidimensional total. Factor meaning, response anchors, measurement invariance, and external performance evidence require further investigation.