Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fair Machine Guidance to Enhance Fair Decision Making in Biased People

Mingzhe Yang, Hiromi Arai, Naomi Yamashita, Yukino Baba

TL;DR

The paper tackles biased human judgments and the scarce guidance on how to adjust decisions toward fairness. It introduces Fair Machine Guidance (FMG), a teacher–student framework that uses iterative machine teaching and fairness-aware learning to educate individuals about their decision criteria and how to modify them for demographic parity. Empirical results show FMG and a bias-feedback baseline both reduce unfairness, but FMG uniquely fosters motivation to reconsider fairness, awareness of bias magnitude, and purposeful changes to decision criteria, albeit with occasional confusion and mixed trust. Overall, the work demonstrates that guiding users through the reasoning and criteria behind decisions—rather than simply presenting final recommendations—can promote more thoughtful, fair judgments in human-AI collaboration.

Abstract

Teaching unbiased decision-making is crucial for addressing biased decision-making in daily life. Although both raising awareness of personal biases and providing guidance on unbiased decision-making are essential, the latter topics remains under-researched. In this study, we developed and evaluated an AI system aimed at educating individuals on making unbiased decisions using fairness-aware machine learning. In a between-subjects experimental design, 99 participants who were prone to bias performed personal assessment tasks. They were divided into two groups: a) those who received AI guidance for fair decision-making before the task and b) those who received no such guidance but were informed of their biases. The results suggest that although several participants doubted the fairness of the AI system, fair machine guidance prompted them to reassess their views regarding fairness, reflect on their biases, and modify their decision-making criteria. Our findings provide insights into the design of AI systems for guiding fair decision-making in humans.

Fair Machine Guidance to Enhance Fair Decision Making in Biased People

TL;DR

The paper tackles biased human judgments and the scarce guidance on how to adjust decisions toward fairness. It introduces Fair Machine Guidance (FMG), a teacher–student framework that uses iterative machine teaching and fairness-aware learning to educate individuals about their decision criteria and how to modify them for demographic parity. Empirical results show FMG and a bias-feedback baseline both reduce unfairness, but FMG uniquely fosters motivation to reconsider fairness, awareness of bias magnitude, and purposeful changes to decision criteria, albeit with occasional confusion and mixed trust. Overall, the work demonstrates that guiding users through the reasoning and criteria behind decisions—rather than simply presenting final recommendations—can promote more thoughtful, fair judgments in human-AI collaboration.

Abstract

Teaching unbiased decision-making is crucial for addressing biased decision-making in daily life. Although both raising awareness of personal biases and providing guidance on unbiased decision-making are essential, the latter topics remains under-researched. In this study, we developed and evaluated an AI system aimed at educating individuals on making unbiased decisions using fairness-aware machine learning. In a between-subjects experimental design, 99 participants who were prone to bias performed personal assessment tasks. They were divided into two groups: a) those who received AI guidance for fair decision-making before the task and b) those who received no such guidance but were informed of their biases. The results suggest that although several participants doubted the fairness of the AI system, fair machine guidance prompted them to reassess their views regarding fairness, reflect on their biases, and modify their decision-making criteria. Our findings provide insights into the design of AI systems for guiding fair decision-making in humans.
Paper Structure (41 sections, 8 figures, 3 tables)

This paper contains 41 sections, 8 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Bias mitigation
  4. Human-AI collaboration for fair decision-making
  5. Human-AI decision making
  6. AI teaching
  7. Fair machine guidance
  8. Overview
  9. Fairness metrics
  10. Fairness-aware ML
  11. Iterative machine teaching
  12. Experiments
  13. Experimental design
  14. Tasks
  15. Participants
  16. ...and 26 more sections

Figures (8)

  • Figure 1: Example of fair machine guidance. (A) Current degree of unfairness (i.e., bias feedback); (B) Teaching materials to teach how to make fair decisions. The materials were selected through iterative machine teaching. The interpretations of the student and teacher models were presented visually.
  • Figure 2: Examples of synthesized profiles
  • Figure 3: Overview of task processing: Initially, participants completed a pre-test and were screened for the next step based on their unfairness scores. Subsequently, they were randomly assigned to each condition (bias feedback or fair machine guidance) and moved to the treatment phase. This phase included the treatment and mini-test, and this cycle was repeated five times. Finally, they underwent a post-test.
  • Figure 4: Scatter plots of unfairness for each participant in the pre- and post-tests. The dotted line represents equal levels of unfairness in both the pre- and post-tests; points below this line indicate an improvement in fairness in the post-test.
  • Figure 5: Distribution of responses to Q10: "Did these tasks cause you to reconsider the fairness of your own decision and that required by society?"
  • ...and 3 more figures