Yoseph Ahmad¹ (), Caeley Tackitt¹, Madison Antkowiak¹, Kaylee Do¹, Sarah Newbolds¹, James Townsend², Michael Wenger¹; ¹The University of Oklahoma, ²Indiana University

This study examines how mnemonic training can improve recognition memory and reduce implicit bias for faces of other races, replicating Lebrecht et al. (2009). The study incorporates EEG measures and an extended training design, and the present report focuses on the behavioral data pertinent to recognition memory and implicit bias. Participants completed a 10-session protocol comprised of pre- and post-assessments and eight training sessions. The baseline and endline sessions included three rounds of a recognition memory task in which participants had to discriminate between studied and unstudied faces. This was followed by the affective lexical priming score (ALPS) task, in which participants observed a prime of a same or other race face, followed by a positive, negative, or neutral word, or a non-word, after which they made their response as to whether the word was a real or a non-word. During training, White participants were assigned to individuate either Black or White faces. Changes in recognition performance and ALPS scores from pre- to post-training were used to assess modulation of the ORE and implicit bias. Results from the recognition memory task showed statistically significant increases in hit rates, decreases in false alarm rates, and increases in recognition discriminability, with no significant change in response bias, replicating the major findings of Lebrecht et al. (2009). Changes in ALP scores showed significant reductions in implicit bias, highlighting the potential for mnemonic training to reduce both the ORE and implicit bias. Ongoing work is intended to determine the extent to which the same effects can be obtained with a perceptual rather than a mnemonic task and the extent to which neural measures are concurrently affected.

Acknowledgements: The University of Oklahoma Vice President for Research and Partnerships