Abstract Background Medical school admission pathways are designed to select suitable applicants, with different approaches potentially impacting students’ learning behaviors and performance. Objective This study aimed to compare students’ self-regulated learning (SRL) strategies, assessment engagement statistics (AES), nongrading evaluation (Outstanding “O”/Satisfactory “S”/Unsatisfactory “U”) preferences, and academic performance across admission pathways, and analyze correlations and linear regression models among summative scores, AES, and course learning outcome (CLO) scores. Methods This mixed methods retrospective observational and cross-sectional survey study used census sampling with selection criteria of all enrolled first-year medical students in 2021 (N=319) across 4 admission pathways: academic (n=23), quota (n=6), test (n=261), and rural (n=29). Demographics included age (19‐24 years) and sex (167/319, 52.4% male). AES, CLO scores, and summative scores were obtained from institutional databases. Two system-embedded institutional questionnaires assessed SRL strategies (316/319, 99.1% response rate) and “O”/“S”/“U” preferences (299/319, 93.7% response rate). Outcome measures included SRL strategies, AES, “O”/“S”/“U” preferences, CLO scores, and summative scores. Statistical significance was set at P <.05. Results When compared among pathways, using one-way ANOVA with Fisher least significant difference post hoc tests, the academic group reported significantly higher mean (with 95% CI) goal setting (4.35, 4.07‐4.63), enthusiasm (4.43, 4.18‐4.69), and lower stress during study (2.64, 2.15‐3.12), while the rural group showed higher pre-examination stress (4.38, 4.10‐4.66) (all P <.05). Most academic (14/22, 63.6%), quota (5/6, 83.3%), and test students (132/243, 54.3%) preferred “O”/ “S”/“U,” while the rural students preferred “S”/“U” (13/28, 46.4%). The academic group showed significantly higher CLO and summative scores but fewer total and intentional attempts and instances of first-pass and highest scoring attempts (all P <.05), whereas the rural group showed significantly lower CLO and summative scores and higher instances of first-pass and highest scoring attempts (all P <.05). For correlation analyses, using Pearson correlation coefficient, summative scores were positively correlated with CLO scores and number of passings and negatively with first-pass attempts. For multiple linear regression analyses, summative scores were positively influenced by number of passings for each CLO and CLO scores and negatively influenced by instances of first-pass attempts and highest scoring attempts. Overall, the academic group demonstrated higher academic performance and fewer attempts and instances of first-pass and highest scoring attempts, while the rural group showed lower academic performance, requiring more attempts for first-passing CLOs. Conclusions Admission pathways significantly influence students’ SRL strategies, AES, evaluation preferences, and academic performance. This study is innovative in analyzing these interconnected components within a single cohort, unlike prior research that examined them separately. By integrating assessment-engagement analytics with SRL data, it offers equity-oriented evidence on how admission systems shape learning behaviors and academic trajectories. These findings provide actionable insights for inclusive curriculum design and early identification of at-risk students. Real-world implications include targeted mentoring, SRL-focused interventions, and assessment reforms balancing academic rigor with psychological safety.
Keadkraichaiwat et al. (2026) studied this question.