# Phase-contingent resilience effects in multilingual medical students: a cross-sectional examination of student demands-resources theory

## TL;DR

This study explores how resilience and language skills affect academic engagement in multilingual medical students across different training phases.

## Contribution

The study introduces phase-contingent resilience effects in medical students, linking them to academic engagement through SDR theory.

## Key findings

- Resilience and Arabic proficiency strongly predict academic engagement in medical students.
- Resilience has a stronger association with engagement in clinical than preclinical students.
- Female students show lower overall engagement compared to male students.

## Abstract

The Student Demands–Resources (SDR) theory proposes that psychological and environmental resources may become more relevant as situational demands increase. This mechanism remains underexamined in medical students. We assessed whether resilience and language proficiency were associated with academic engagement across training phases in a multilingual medical cohort.

We surveyed 422 medical students at an international medical school (May–September 2024). Hierarchical multiple regression modeled academic engagement (UWES-9S) as a function of psychological resilience (BRS), Arabic language proficiency, clinical training phase, prior residence, and social support (DSSI). The Resilience × Clinical Phase interaction was specified a priori; other interactions were exploratory.

The main-effects model explained 52.8% of variance in engagement (R2 = 0.528). Resilience (β = 0.418, f2 = 0.370), sex (β = −0.410, f2 = 0.356), and Arabic proficiency (β = 0.370, f2 = 0.267) showed the largest effects. The resilience–engagement association was stronger in clinical than preclinical students (standardized β: 0.630 vs. 0.291; interaction β = 0.339, p < 0.001, f2 = 0.129). Female students showed lower overall engagement, with differences concentrated in specific dimensions. Social support showed small positive associations across dimensions. Sensitivity analysis excluding prior residence yielded near-identical estimates and unchanged inference (reduced-model R2 = 0.525; ΔR2 = −0.003).

These hypothesis-generating findings are consistent with phase-contingent resource–engagement associations in SDR theory within medical education. The cross-sectional design and indirect demand proxy do not support causal inference.