Bayesian multiple membership multiple classification logistic regression model on student performance with random...

Area of Science:

• Educational Research
• Higher Education Studies
• Academic Analytics

Background:

• Student retention and graduation are critical metrics for academic institutions.
• Identifying factors influencing student academic performance is essential for institutional improvement.
• Defining academic success involves achieving specific Grade Point Average (GPA) thresholds.

Purpose of the Study:

• To identify key factors contributing to college student academic success.
• To analyze the impact of instructors and academic majors on student performance.
• To develop analytical models for understanding student retention and graduation.

Main Methods:

• Utilized a large dataset from a state university spanning three semesters.
• Employed multiple membership multiple classification (MMMC) Bayesian logistic regression models.
• Incorporated random effects for instructors and majors to account for nesting and complexity.

Main Results:

• Bayesian models identified significant factors influencing college student academic performance.
• Instructors and academic majors were found to account for a substantial portion of student academic success.
• Models adjusted for variables including residency status, academic level, and student athlete status.

Conclusions:

• Instructors and majors are significant indicators of student retention and graduation rates.
• These factors are integral to university recruitment and student competition strategies.
• The study highlights the importance of faculty and departmental influence on educational outcomes.