Effectiveness of learner-centered learning in clinical medical courses: a mixed-methods study

Learner-centered learning (LCL) has gained attention in medical education, yet robust evidence comparing its effectiveness with traditional teaching (TT) remains limited. Our pilot study found no significant difference in knowledge scores but suggested a possible trend toward greater teaching stability. This study expanded the sample size to further evaluate LCL. In this mixed-methods study, 27 medical students generated 135 course-student records (58 TT, 77 LCL). The primary analysis was a linear mixed model with student ID as a random intercept. Independent-samples t-tests and chi-square tests were used only as simple unadjusted comparisons. Open-ended responses were thematically coded, and a sensitivity analysis excluded three students who participated in only one teaching method. Ethical approval was obtained from Shanghai Pudong Hospital, and all participants provided informed consent. In the primary mixed-model analysis, no significant difference was observed between LCL and TT in knowledge scores (β = 0.022, p = 0.922). t-Tests showed similar unadjusted results for knowledge (19.31 ± 1.50 vs. 19.36 ± 1.31, p = 0.826), single-choice (9.86 ± 0.44 vs. 9.83 ± 0.55, p = 0.725), and multiple-choice scores (9.45 ± 1.17 vs. 9.53 ± 1.02, p = 0.657). LCL showed lower score variability (SD: 1.307 vs. 1.501; variance ratio: 1.32), suggesting a trend toward more consistent outcomes. LCL courses achieved perfect satisfaction (5.00 ± 0.00); because questionnaire dimensions differed, satisfaction was not compared. LCL students mentioned mind maps more often (22.1% vs. 8.6%) and had more positive feedback (70.1% vs. 58.6%; p = 0.112). Theme frequencies are descriptive due to prompt differences. Sensitivity analysis was consistent with the main analysis. No statistically significant difference was found between LCL and TT in knowledge outcomes. However, given the limited sample size and power, this should not be interpreted as evidence of equivalence or non-inferiority. Larger studies are needed to determine whether LCL is a viable alternative. The potential benefits of LCL for teaching consistency and student satisfaction warrant further investigation.