Our symposium presents three projects that use machine learning to generate theoretical insights about prejudice in organizations. The first project illustrates the uses of natural language processing to examine how commonly used AI-based HR hiring tools are biased against applicants who sporadically use African American English instead of Standard American English. The following two projects use machine learning to identify antecedents of racism and xenophobia through abductive reasoning. Both studies identified novel antecedents that were not predicted by existing theories, and verified the antecedents suggested by machine learning with multiple correlational, experimental, and archival studies. These insights can subsequently be used to develop interventions to reduce racism and xenophobia in the workplace. You Don’t Sound Black: Codeswitching, AI, and the Use of HR Platforms in Candidate Selection Presenter: Kofi Arhin; Lally School of Management, Rensselaer Polytechnic Institute Presenter: Louis Hickman; The Wharton School, U. of Pennsylvania Presenter: Jason Nicholas Kuruzovich; Rensselaer Polytechnic Institute Pride and Prejudice: Pride in National History Increases Racism Presenter: Elizabeth Demissie Degefe; Nanyang Technological U. Presenter: Abhishek Sheetal; Department of Management and Marketing, The Hong Kong Polytechnic U. Belief in Free Choice Leads to Pro-Immigration Attitudes Presenter: Kevin Nanakdewa; U. of Toronto
Read full abstract