Standardizing College GPA for University Senior Intake Admissions: A Moderation Tool for Equitable Evaluation

Abstract

University admissions follow two distinct pathways: first-year admissions using standardized tests (e.g., SAT in the United States, GaoKao in China) and senior intake admissions, where students enter university based on their college GPA. While first-year admissions benefit from uniform comparison metrics, senior intake processes rely on college GPAs that vary significantly across institutions. This paper addresses this standardization gap by developing and validating methods to calibrate disparate college GPAs. Recognizing that raw GPA is a biased measure influenced by institutional grading policies, we propose three moderation models to align raw College GPAs (X) with observed University Performance (Z): (1) a Mean-Adjusted model, (2) Direct Benchmark-to-Raw Score Regression, and (3) Inter-Score Regression with Benchmark Standardization. Evaluation using synthetic datasets (training N=75; test sets N=95, N=29) demonstrates that Inter-Score Regression with Benchmark Standardization produces the most substantial improvement in predictive validity (ΔR² = +0.40 for training, +0.24 and +0.26 for test sets), maintaining robustness across varying sample sizes and grade distributions. This research provides admissions officers with a standardized evaluation tool for senior intake, advancing equitable assessment practices in higher education.