Measuring Non-Expert Comprehension of Machine Learning Fairness Metrics

Debjani Saha, Candice Schumann, Duncan Mcelfresh, John Dickerson, Michelle Mazurek, Michael Tschantz
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:8377-8387, 2020.

Abstract

Bias in machine learning has manifested injustice in several areas, such as medicine, hiring, and criminal justice. In response, computer scientists have developed myriad definitions of fairness to correct this bias in fielded algorithms. While some definitions are based on established legal and ethical norms, others are largely mathematical. It is unclear whether the general public agrees with these fairness definitions, and perhaps more importantly, whether they understand these definitions. We take initial steps toward bridging this gap between ML researchers and the public, by addressing the question: does a lay audience understand a basic definition of ML fairness? We develop a metric to measure comprehension of three such definitions–demographic parity, equal opportunity, and equalized odds. We evaluate this metric using an online survey, and investigate the relationship between comprehension and sentiment, demographics, and the definition itself.

Cite this Paper


BibTeX
@InProceedings{pmlr-v119-saha20c, title = {Measuring Non-Expert Comprehension of Machine Learning Fairness Metrics}, author = {Saha, Debjani and Schumann, Candice and Mcelfresh, Duncan and Dickerson, John and Mazurek, Michelle and Tschantz, Michael}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {8377--8387}, year = {2020}, editor = {Hal Daumé III and Aarti Singh}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/saha20c/saha20c.pdf}, url = { http://proceedings.mlr.press/v119/saha20c.html }, abstract = {Bias in machine learning has manifested injustice in several areas, such as medicine, hiring, and criminal justice. In response, computer scientists have developed myriad definitions of fairness to correct this bias in fielded algorithms. While some definitions are based on established legal and ethical norms, others are largely mathematical. It is unclear whether the general public agrees with these fairness definitions, and perhaps more importantly, whether they understand these definitions. We take initial steps toward bridging this gap between ML researchers and the public, by addressing the question: does a lay audience understand a basic definition of ML fairness? We develop a metric to measure comprehension of three such definitions–demographic parity, equal opportunity, and equalized odds. We evaluate this metric using an online survey, and investigate the relationship between comprehension and sentiment, demographics, and the definition itself.} }
Endnote
%0 Conference Paper %T Measuring Non-Expert Comprehension of Machine Learning Fairness Metrics %A Debjani Saha %A Candice Schumann %A Duncan Mcelfresh %A John Dickerson %A Michelle Mazurek %A Michael Tschantz %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-saha20c %I PMLR %P 8377--8387 %U http://proceedings.mlr.press/v119/saha20c.html %V 119 %X Bias in machine learning has manifested injustice in several areas, such as medicine, hiring, and criminal justice. In response, computer scientists have developed myriad definitions of fairness to correct this bias in fielded algorithms. While some definitions are based on established legal and ethical norms, others are largely mathematical. It is unclear whether the general public agrees with these fairness definitions, and perhaps more importantly, whether they understand these definitions. We take initial steps toward bridging this gap between ML researchers and the public, by addressing the question: does a lay audience understand a basic definition of ML fairness? We develop a metric to measure comprehension of three such definitions–demographic parity, equal opportunity, and equalized odds. We evaluate this metric using an online survey, and investigate the relationship between comprehension and sentiment, demographics, and the definition itself.
APA
Saha, D., Schumann, C., Mcelfresh, D., Dickerson, J., Mazurek, M. & Tschantz, M.. (2020). Measuring Non-Expert Comprehension of Machine Learning Fairness Metrics. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:8377-8387 Available from http://proceedings.mlr.press/v119/saha20c.html .

Related Material