Degree Date

3-2025

Document Type

Dissertation - Public Access

Degree Name

DBA Doctorate in Business Administration

Academic Discipline

Business Administration

First Advisor

Dr. Marguerite Chabau

Second Advisor

Dr. Colleen Ramos

Third Advisor

Dr. David SanFilippo

Abstract

This quantitative study closely emulated Zou and Khern's (2022) analysis of AI Bias in Mortgage Applications. They used the Home Mortgage Disclosure Act (HMDA) dataset from the Federal Financial Institution Examination Council's (FFEIC) website to review mortgage loan data from 2019 to determine if there was bias in the AI mortgage application approvals. Those researchers concluded that bias does exist in the AI mortgage underwriting software. Since their study, research has demonstrated that discrimination continues to exist in AI software. Therefore, this research expands on Zou and Khern’s study to determine if there are differences in the mortgage loan approval outcomes, whether AI Bias is present in the mortgage application approvals, and if fair AI algorithms reduce AI bias in the mortgage application datasets by analyzing historical mortgage loan data from the HMDA dataset published in 2022. The variables in this study included race and region as an independent variable and mortgage loan approval outcome as a dependent variable. The statistical analysis included a Chi-square test to analyze the relationship between race, geographical, and loan approval outcomes. The methodology included the fair-on-average causal Effect (FACE) and fair-on-average causal Effect on the Treated (FACT) to detect AI bias in the dataset. Additionally, IBM AI Fairness 360 (AIF360) and Microsoft Fairlean (MSF) were used to detect and mitigate bias. The findings concluded that bias does exist in the mortgage application dataset. The research highlighted the need for fair AI algorithms to reduce bias in the mortgage approval process.

View Full Text

Share

COinS