Preview
Creation Date
Spring 2026
Description
My research is studying listener evaluation of a speaker based on the speaker’s dialect. I hypothesize that listeners will have negative evaluations of speakers with unfamiliar dialects and that women will receive more strongly negative evaluations than men. The research has implications for anti-discrimination policy and AI development and will include New England, Southern White, Midland, and Western dialects with special attention being given to uncovering covert listener bias. The two masked listeners on either side of the female speaker in my image have stoic faces, representing how covert bias is not obvious on the outside; their “real” emotions are on the paper faces they hold. Behind the speaker is a man, arm aloft and thumb to the side waiting to deliver judgment; this illustrates the power others hold over speakers when they discriminate for or against them.
Photo Credit: Used with permission from Emmanuel Owusu.