Abstract: Objective To analyze the students' learning situation, improve the quality of the test papers, and provide reference for further teaching reform, all based on the test results of medical statistics in “5+3” program clinical majors.Methods The principles and methods of educational measurement are used to analyze the difficulty, discrimination,reliability (Cronbach α coefficient) and coverage and test scores of the test paper. Results The average test score of 147 “5+3” program medical students was 78.25±11.70, ranged from 42.5 to 98.5, and the passing rate was 91.2%. The average scoring rates for choice questions, terms explanations, fill-in-the-blank questions, short-answer questions, and calculation questions were 85%, 87.2%, 78%, 74.1%, and 76.3%, respectively. The difficulty coefficient of the test paper was 0.22, and the discrimination index was 0.28. The Cronbach α coefficient of the test paper was 0.846, and the correlation coefficient between the questions less than 0.2 accounted for 89%. Conclusion The students in this test had excellent grades; the difficulty of the test paper was slightly lower, the discrimination index was large, the reliability was common, and the test coverage was wide. The quality of this test paper was good, which could reflect the level of knowledge of students. In the future teaching, the operation of multi-statistical knowledge and statistical software should be carried out to improve the scientific research quality of students.

Key words: clinical medicine, “5+3” program majors, medical statistics, test paper quality, teaching reform, statistical software