Analysis of factors contributing to student’s academic performance in college’s STEM program

The purpose of this study is to analyze different variables in understanding how they affect the academic performance of students taking STEM courses at a mid-sized college in the southern United States. These factors are classified into pre-college variables and college variables. The pre-college variables are ACT/SAT scores, gender, ethnicity, and first-generation student or not. The college variables are college GPA, Pell grant recipient, the student’s continuance with the degree major in sophomore year. Implementing regression analysis and using backward regression to remove variables that are not correlated to the response variable (CGPA), we were able to arrive at the best model using the three variables: ACT mathematics, ACT English, and gender. The resulting variables selected show us how the value of R^2 has helped us in making sense of the model selected by 14.2%.