Creating an Advanced Statistics Interest Group to Increase Medical Students' Confidence and Competence in Statistics and Data Analysis.

INTRODUCTION: Post-graduation exit surveys have consistently demonstrated that the subjects of statistics and data analysis are considered a challenging aspect of medical education. This is reflected in student attitudes towards the subjects, which are often neutral to negative. Whether due to a lack of exposure to the subject or decreased retention of their education, practicing physicians often lack advanced statistics and data analysis skills. As data analysis continues to become ever more important for the practice of evidence-based medicine, many medical school curricula only provided a limited education on the subject. Previous studies have shown that the implementation of a student-taught supplementary education in statistics and data analysis can improve students' attitudes toward statistics and their confidence in performing data analysis.

METHODS: For this project, an advanced statistics and data analysis interest group was created, called the Fisher Society. This organization provided an additional two-year curriculum in advanced statistics and data analysis to medical students. The organization met bimonthly with meetings consisting of a lecture followed by deliberate practice using the discussed methods of data analysis. The Fisher Society also hosted an annual public lecture on high-yield statistics and data analysis topics covered on medical licensure and board exams. In conjunction with the student interest group, an annual survey (M-STATS) was sent to the entire Sanford School of Medicine medical student body that was used to gauge medical students' perceived competence and confidence in the subjects of statistics and data analysis. The survey consisted of 11 questions assessing students' confidence, competency, and exposure to statistics and data analysis using the five-point Likert scale.

RESULTS: Between 2021 and 2022, no statistically significant difference was observed in the average response to the questions on the M-STATS within the entire medical student cohort. Data from 1st and 2nd year medical students demonstrated a statistically significant decrease in students' perceived difficulty with statistics from 2021 to 2022. Additionally, a statistically significant increase in student-perceived knowledge of statistics and data analysis was displayed in the data from 1st and 2nd year students. Between 2021 and 2022, a statistically significant increase in confidence for 1st and 2nd year medical students in teaching data analysis and statistics was also observed.

CONCLUSIONS: Between 2021 and 2022, there was no statistically significant change in the confidence or competence in the fields of statistics and data analysis among the medical students at the Sanford School of Medicine. However, promising trends are beginning to show increases in positive attitudes towards the subjects as well as decreases in perceived difficulty of the subject. Additionally, the 1st and 2nd year medical student responses to the 2022 survey displayed greater confidence and decreased perceived difficulty with statistics and data analysis when compared with responses from the 2021 survey.