Abstract: Objective To analyze medical undergraduates' preferences on MOOC learning programs, in order to inform MOOC developers and managers to further optimize MOOC resources in medicine. Methods The discrete choice experiment method was used to determine the potential attributes and levels affecting medical undergraduates' MOOC learning preferences through literature review, student interviews, expert consultation, and importance ranking. SAS 9.4 and Stata 17 were used to calculate the optimal choice set numbers and design the discrete choice experiment questionnaire. A questionnaire survey was conducted among 453 full-time medical undergraduates from a higher medical college in Chongqing, and the conditional logit model was used to quantify the preference degree of medical undergraduates for MOOC learning attributes. Results This study collected 375 valid questionnaires, and six attributes including course type, content, difficulty level, duration, summary method, and correlation with formative assessment scores had a significant impact on medical undergraduates' preferences for MOOC learning programs (all P<0.05). The preference for correlation with formative assessment scores (no correlation) was most significant (β=0.38), with the highest relative importance score (0.301). Assuming the removal of correlation with formative assessment scores (no correlation) from the optimal scenario, the probability of students choosing this MOOC learning plan will decrease by 6.14%; compared with public basic courses, the preference for experimental courses (β=0.34) was significantly higher than that for theoretical courses (β=0.16); for the lower-grade group (first and second year) and the higher-grade group (third year and above), the most valued attribute levels were the course summary method (comprehensive lecture of knowledge points) (β=0.35, P<0.001) and course duration (<10 minutes) (β=0.45, P<0.001), respectively. Conclusions Medical undergraduates do not want MOOCs to be correlated with formative assessment scores and prefer to choose experimental MOOCs for learning. There is heterogeneity in the preferences for MOOC learning programs among medical undergraduates at different learning stages. It is recommended to reduce the correlation between medical undergraduates' MOOC scores and formative assessment scores in blended teaching; to focus on supplementing, updating, and integrating medical experimental MOOCs, to explore a new model of full online experimental teaching combining MOOCs and virtual simulation experiments; to emphasize ″learning according to needs,″ and to build a MOOC teaching system suitable for medical undergraduates at different learning stages to improve the quality and efficiency of MOOC learning.

Key words: Students, medical, MOOCs, Discrete choice experiment, Learning program, Preference