Polish translation and validation of the student satisfaction and self-confidence in learning scale among healthcare students at the Pomeranian Medical University in Szczecin: a single-center cross-sectional validation study.

BACKGROUND

Simulation-based education is an established component of healthcare training, yet robust Polish-language instruments for evaluating the learner's reaction to simulation remain scarce. The Student Satisfaction and Self-Confidence in Learning (SSCL) scale operationalizes the reaction level of educational evaluation within the NLN/Jeffries Simulation Framework. This study aimed to translate, culturally adapt, and validate the Polish version (PL-SZZ) in a single-center, convenience sample of healthcare students and, secondarily, to determine how it should be scored in practice.

METHODS

This cross-sectional validation study was conducted at the Pomeranian Medical University in Szczecin (data collected 2023-2025). Translation followed a five-stage forward-backward protocol with sworn translators and expert adjudication. Students who had participated in simulation classes completed the PL-SZZ anonymously. After screening for extreme observations (1.5 × interquartile range), 827 of 879 questionnaires were analysed. Confirmatory factor analysis (DWLS/WLSMV applied to polychoric correlations) compared a one-factor and a correlated two-factor model. Reliability was evaluated using Cronbach's α, McDonald's ω, and marginal reliability (MR); validity via standardized loadings, average variance extracted (AVE), the heterotrait-monotrait (HTMT) ratio, and added-value analysis.

RESULTS

The analytic sample comprised 827 students (86.6% female), predominantly nursing students (64.6%). The correlated two-factor model fitted better than the unidimensional model (χ2(101) = 401.89; CFI = 0.986, TLI = 0.989, RMSEA = 0.060, SRMR = 0.050). Standardized loadings ranged from 0.701 to 0.886 across items, and the two factors were strongly correlated (r = 0.860; HTMT = 0.866). Internal consistency was good (total α = ω = 0.909; subscales 0.838-0.871), whereas marginal reliability was lower (0.656-0.728); added-value coefficients were negative for both subscales (-0.071 and -0.241). Scores did not differ across simulation modalities but differed significantly between fields of study and by sex.

CONCLUSION

In this single-center, predominantly female and nursing-dominated sample, the PL-SZZ reproduced the original correlated two-factor structure with acceptable reliability and construct validity. Because the subscales added no incremental precision, the total score is recommended for routine evaluation, with subscale scores reserved for diagnostic and curriculum-development purposes. Findings should not be generalized beyond comparable populations. Further multi-center research is needed before using the instrument for between-group or institutional comparisons.