High student attrition rate has been an ever prevalent issue for educational institutions for many years. In the STEM field specifically, student attrition rate is proved to be higher than other fields of study. This problem is growing with the changes brought up by the COVID-19 pandemic which makes it more difficult for many students to stay in the institutions. This study utilized Survival Analysis as a statistical tool to predict the student attrition of Engineering undergraduates at Adamson University, and to identify the contributing factors that involve it. Results from the Cox Proportional Hazard Regression indicated that out of 22 variables, religion, location, parent’s educational attainment, information and technology factor, student engagement, COVID-19 related factor and university environment condition were found to be significant factors and predictors of turnover. The Survival and Hazard Function were also estimated. It was found that the presence of time as a variable can help in calculating the probability until attrition happens. Survival Analysis was also used to see the trend of attrition and the students who’s more likely to experience attrition, through the use of survival and hazard functions. The Kaplan-Meier estimator was used to calculate the probability of a student to survive at a certain time interval. This study opened a new avenue for researchers to use the survival analysis to predict the graduation time, and scrutinize the factors for graduation delay through the inclusion of the 5th year respondents in the study.

Keywords - survival analysis, student attrition, cox regression, Kaplan-Meier method