Comparing traditional and online problem-based learning in child and adolescent psychiatry in Nagoya: a novel statistical approach in Japanese educational settings.

backgroundProblem-Based Learning (PBL) requires active communication and learner autonomy, which can be challenging in cultural contexts that emphasize group harmony and indirect communication, such as Japan. This challenge is often amplified when PBL is conducted in a non-native language (English), potentially inducing "foreign language anxiety." While the COVID-19 pandemic necessitated a shift to online learning, the specific impact of this format on psychological barriers and student engagement in high-context cultures remains underexplored. We investigated whether the online format could reduce these transactional distances and enhance learning outcomes compared to traditional in-person PBL.

methodsWe conducted a naturalistic, historical control study comparing fourth-year medical students at Nagoya University during a Child and Adolescent Psychiatry curriculum. The 2019 cohort (n = 109) participated in-person, while the 2021 cohort (n = 100) participated online. We administered a 15-item questionnaire assessing satisfaction, engagement, and case suitability. Beyond standard descriptive comparisons, we applied a novel multidimensional statistical framework. This included Ordered Logistic Regression adjusted for age and sex to identify predictors of satisfaction, Exploratory Factor Analysis (EFA) to validate the instrument's latent structure, and K-means Clustering to identify distinct student "phenotypes" based on response patterns.

resultsThe online cohort reported significantly higher satisfaction across most domains (Odds Ratios 0.36-0.51, p < 0.05). Factor analysis identified two primary dimensions-"Perceived Learning Efficacy" and "Engagement"-while "Communication Skills" (Question 4) emerged as an independent outlier, failing to load on either dimension. Cluster analysis identified two distinct student phenotypes: a "Traditional/Dissatisfied" profile (Cluster 1: n = 77, 69% from the in-person group, predominantly male and older) and a "Digital/Satisfied" profile (Cluster 2: n = 127, 57% from the online group, predominantly female and younger).

conclusionsOnline delivery of PBL was associated with significantly higher student satisfaction and engagement scores in this Japanese context. Because the study is observational (historical-cohort comparison with no randomisation), this association cannot be attributed causally to the modality, and the role of reduced foreign-language anxiety is an interpretive hypothesis that the present design does not directly test. The identification of distinct student phenotypes suggests that demographic factors (gender, age) and delivery modality interact to shape the learning experience. These findings advocate for a tailored pedagogical approach, where digital formats serve as a "safe harbor" for students with high communication apprehension.