Faculty Compensation V.S. Student Success

Exploring how faculty compensation influences student outcomes (GPA and retention rates) at Purdue University to optimize budget allocation.

role

Project Lead

aim

Model a budget for Purdue University based on faculty compensation

duration

January 2025 - May 2025

problem

Universities invest heavily in faculty salaries, but the impact on student success remains unclear.

  • Administrators need data-driven insights to optimize budget allocation for better student outcomes.

  • Faculty salaries may influence GPA, retention, and graduation rates, but the relationship is not well understood.

  • Decision-makers require statistical analysis to determine if higher compensation leads to improved student performance.

Key Questions

  1. Does faculty compensation impact student success?

  2. How can universities optimize faculty salaries to enhance student performance?

  • Data Sources

    • Purdue Data Digest (Budget Allocation, Retention Rates)

    • BoilerGrades.com (Course Grades)

    • Indiana Gateway (Faculty Salaries)

  • Data Summary

    • Budget: Covers nine academic years (2015-2024).

    • Course Grades: Over 78,000 course records.

    • Salaries: Over 247,000 salary records, filtered to focus on faculty.

    • Retention: 11 years of retention data.

data

data preparation

Data Cleaning

  • Standardized column names and formats across datasets.

  • Filtered salary data to include only relevant faculty (ignored TA’s and assistant faculty, for instance).

  • Created derived attributes like "Retention Rate" and "Average Salary per Department."

  • Removed unnecessary columns (e.g., addresses in salary data).

  • Standardized grading by combining A+, A, and A- into a single "A" category.

  • Merged faculty salaries with course grades based on instructor names.

  • The Salary & Wages category has consistently grown over the years, highlighting its increasing share of Purdue’s budget.

  • Some instructors consistently give higher or lower grades, indicating different grading standards or course difficulty.

  • Faculty salaries vary significantly by department, with some departments receiving notably higher compensation.

  • The Left with No Degree category increased by a little every year, which could indicate a relationship between difficulty of classes and students leaving.

patterns in the data

model

We developed a hierarchical regression model that considers the faculty, department, and student levels. This ensures a data-driven approach to optimizing university spending.

  • Faculty Grading Behavior (GradesGiven): Measures whether lenient or strict grading impacts retention.

  • Department Budget Allocation (DeptBudget): Evaluates whether well-funded departments have higher retention rates.

  • Random Effects (u_j & v_k): Accounts for unobserved variations at faculty and department levels.

  • Student-Level Variation (ε_{ijk}): Captures individual student differences.