Predicting the Possibility of Student Admission into Graduate Admission by Regression Model: A Statistical Analysis

Abstract

This study aims to alleviate the uncertainties faced by prospective students during the application process by developing a predictive model for admission probabilities based on CGPA and GRE scores. The research investigates the significance of these predictor variables about the response variable, "Chance of Admit." Employing linear regression analysis, the model is thoroughly examined to evaluate its adequacy, predictive accuracy, and the need for interaction terms. The findings indicate that both CGPA and GRE scores play a crucial role in forecasting admission chances, with an adjusted R2 value of 0.0835, suggesting an 80% reduction in variance around the regression compared to the main line. The diagnostic plot of the model confirms its precision, revealing minimal deviations from linearity and normality in residuals. Furthermore, the study addresses concerns about multicollinearity using the Variable Inflation Factor (VIF) and finds no significant correlation between GRE Scores and CGPA. In summary, this research presents a robust predictive model for student admission probabilities, offering valuable insights for both prospective applicants and educational institutions.