Tracie Stewart (Georgia State University) will use this award from Russell Sage Foundation to explore how to reduce implicit stereotypes and bias between groups through a technique she developed called “situational attribution training.” This technique attempts to shift people away from using stereotyped explanations of the behavior of others and more toward an appreciation of situational causes. Repeated exposure to situational attribution training has shown a reduction in automatic stereotyping by helping to establish new cognitive habits that do not rely on stereotypes. All of Stewart’s studies will use white American students of Georgia State University recruited through the psychology department. In order to measure the lasting potential of this training, half of 120 participants will take an assessment immediately after the training and half will return to the lab for an assessment twenty-four hours later. In addition, this round of experiments will include a recording of facial EMG responses which are a stronger predictor of actual racial discrimination than the more common implicit bias measures.
A second study of 240 white American college students will examine whether the effects of this training can be generalized across both high and low cognitive demands by limiting participants’ cognitive resources. This experiment is identical to the first study with the addition of a recall task of a 1-digit (low load) up to 5-digit (high load) number at the end of each trial. The experiment will also employ a pre-screening session to measure the strength of each participant’s implicit and explicit racial stereotyping and prejudice. The investigator will then ascertain whether attribution training is equally effective for participants with high versus low initial levels of stereotyping and prejudice, or whether highly biased participants might be less impacted by this training.
A final study will bring this training into a simulated real-world situation to see if the attribution training technique will function in a “shoot/don’t shoot” paradigm. Participants will be given the attribution training as described in earlier experiments. Following the training, they will complete a computer trial with images of perpetrators in a virtual inner-city setting and will be asked to indicate a decision to shoot or not shoot. The predicted finding for this study is that participants will be slower to shoot unarmed African American perpetrators following attribution training due to a reduction in their automatic negative stereotyping relative to the control participants who did not receive the training.