This software tool uses unbiased, comprehensive dissimilarity analysis on a pool of qualified applicants to achieve optimal holistic diversity among accepted applicants. The use of affirmative action policies in recruiting, hiring, or enrolling new employees or students is a controversial issue that has gone through a series of legal developments. Recent U.S. Supreme Court decisions and state legislative acts have concluded that attributes such as race, gender, color, or national origin may not receive preferential treatment in determining an applicant’s admittance. Consistent with these rulings, however, are holistic review systems that admit a diverse class of applicants without giving preference to any single attribute over another. Although holistic review by committee is a widespread practice among academic institutions and employers, the process is uniquely vulnerable to legal challenges due to inherent subjectivity. Since human analysis plays a major role in the selection of applicants under these processes, it is possible that review committee members fail to equally consider all of an applicant’s attributes, resulting in a potentially biased evaluation of that applicant or other applicants. Researchers at the University of Florida have designed a human-centered data-mining tool that analyzes pools of applications to optimize the holistic diversity of the group of accepted applicants and remove bias in admissions and recruiting. By evaluating the qualifying applications based on their relative differences and similarities with respect to the entire pool, this software will achieve a greater level of diversity among those selected without giving any specific attribute preferential treatment.
Application data analysis software package that removes human bias in employment recruitment and in the admission process of post-secondary institutions, while achieving maximum holistic diversity and academic achievement levels
This software platform optimizes the holistic diversity of accepted applicants. Admissions and recruiting teams indicate specific necessary qualifications in the platform, and applicants that do not meet these qualifications immediately filter out of the pool, while those that do pass on to the next stage and are considered qualified. The admissions committee indicates the number of available slots or offers. After comparing each remaining application with every other, according to their various nominal and numeric attributes, the platform then groups the applications into clusters and recommends an applicant from each cluster. Since these clusters represent applications that are holistically similar, the group of recommended applicants will be holistically diverse. This approach eliminates bias in admissions and hiring decisions while empowering admissions and hiring officials to achieve qualified, diversity within the limits of the law.