Professors Alan Mislove and Christo Wilson wanted to test a number of housing and employment website algorithms for the presence of hidden discrimination. While algorithms do not have any predisposition against any group, faulty programming can create deplorable discriminatory effects.2 Without testing, it can be difficult to tell which algorithms discriminate and on what grounds. Mislove and Wilson devised an "audit testing" model where researchers designed the profiles of two groups of paired individuals that are equally qualified in the market being studied. Under the "audit testing" model, one group consists of legally-protected minority members, and the other consists of individuals without those minority characteristics.
Brian Z. Mund,
Protecting Deceptive Academic Research under the Computer Fraud and Abuse Act,
Yale L. & Pol'y Rev.
Available at: https://digitalcommons.law.yale.edu/ylpr/vol37/iss1/7