156 - Can We Predict Your Performance? Exploring Relationships Between Pre-matriculation Data and Gross Anatomy Performance

Poster Board Number: 156
There are separate poster presentation times for odd and even posters.
Odd poster #s – first hour
Even poster #s – second hour

Co-authors:

Nancy Adams - University of Florida; Phoung Huynh - University of Florida; Kyle Rarey - University of Florida

Abstract Body : Introduction and Objective: Educational data mining (EDM) and predictive analytics in medical education has been justified to assist admissions committees to choose students that will be most likely to matriculate and perform well and assist student support personnel with identifying at-risk students for the purpose of initiating interventions. This study's purpose is to see if the medical school entry metrics could predict first semester anatomy performance. Methods: De-identified pre-admissions data from one cohort of 133 students, containing continuous predictors: MCAT scores, cumulative science GPA, graduate bcpm GPA, total GPA, SES value, age, and experience count; as well as categorical predictors: science vs. non-science undergraduate degree, highest degree earned, first-generation status, and Underrepresented in Medicine (URM). The anatomy lab practical scores (three total) were averaged together to form one output- a continuous outcome and were also coded into a binary categorical outcome- pass all exams or fail at least one. Block entry multiple regression analysis of the predictors on the average anatomy lab practical score was conducted. A logistic regression model was also developed to explore the relationship between the categorical outcome and two predictor variables: MCAT score and first-generation status. Results: The results of the block entry regression analysis showed that continuous predictors, cumulative science GPA, total GPA, and MCAT total scores are positive and statistically significant predictors of anatomy performance; and that categorical predictors, SES value and First generation are significant negative predictors of academic performance on the lab practicals. The logistic regression analysis result showed that being a first-generation college student is a negative but not statistically significant predictor of passing the lab practicals. However, the odds ratio for being a first-gen student shows that for the same score on the MCAT, the odds of passing the practicals decrease 53%. Conclusions: The results showed that Cumulative science gpa, total gpa, and MCAT scores, and First-generation predictors have some correlation with gross anatomy performance. These predictors could serve as potential markers for performance in gross anatomy. Limitations: A major limitation of this type of study is that there is no data collected for students that did not get accepted to the medical school cohort. Significance and Implications: The long-term goal would be to utilize the formulated regression model to encourage educational practice makers within medical education to consider programs and activities that assist in student development and at-risk students, such as anatomy bootcamps before matriculating into medical school.