Skip to main content
. 2026 Oct 9;14:1958746. doi: 10.3389/fpubh.2026.1958746

Table 2.

Pre- and post-course outcomes and exploratory baseline-adjusted comparisons.

Outcome Time Intervention class (n = 65), mean ± SD Control class (n = 75), mean ± SD Adjusted mean difference (95% CI) F(1,137) Partial η2 p Holm-adjusted p
Self-directed learning Pre 65.89 ± 12.91 68.99 ± 13.72
Post 74.48 ± 14.79 66.95 ± 11.20 7.68 (3.29–12.07) 11.96 0.08 <0.001 0.009
Critical thinking Pre 108.98 ± 16.28 107.73 ± 23.20
Post 128.29 ± 11.36 113.00 ± 10.87 15.27 (11.53–19.00) 65.41 0.323 <0.001 <0.001
Learning engagement Pre 62.37 ± 17.99 59.65 ± 13.49
Post 69.65 ± 17.94 62.69 ± 13.09 7.06 (1.82–12.30) 7.1 0.049 0.009 0.034

ANCOVA models included the post-course score as the dependent variable, study condition as the factor, and the corresponding pre-course score as the covariate. Positive adjusted differences favor the intervention class. Holm adjustment was applied across the 17 comparisons reported in Tables 2–4. Because one intact class was included in each study condition, the individual-level analyses were exploratory and did not account for between-class heterogeneity. ANCOVA, analysis of covariance; CI, confidence interval; SD, standard deviation.