Date of Award

Summer 8-18-2010

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Tracie L. Stewart


Individuals tend to underestimate situational causes and overly rely on trait causes in explaining negative behaviors of outgroup members, a tendency named the ultimate attribution error (Pettigrew, 1979). This attributional pattern is directly related to stereotyping, because attributing negative behaviors to internal, stable causes tends to perpetuate negative stereotypes of outgroup members. Recent research on implicit bias reduction revealed that circumventing individuals’ tendency to engage in the ultimate attribution error led to reduced stereotyping. More specifically, training White participants to consider situational factors in determining Blacks’ negative stereotypic behaviors led to decreased automatic stereotype activation. This technique was named Situational Attribution Training (Stewart, Latu, Kawakami, & Myers, 2010). In the current studies, I investigated the mechanisms and moderators of Situational Attribution Training. In Study 1, I investigated the effect of training on spontaneous situational inferences. Findings revealed that training did not increase spontaneous situational inferences: both training and control participants showed evidence of spontaneous situational inferences. In Study 2, I investigated whether correcting trait inferences by taking into account situational factors has become automatic after training. In addition, explicit prejudice, motivations to control prejudice, and cognitive complexity variables (need for cognition, personal need for structure) were investigated as moderators of training success. These findings revealed that Situational Attribution Training works best for individuals high in need for cognition, under conditions of no cognitive load, but not high cognitive load. Training increased implicit bias for individuals high in modern racism, regardless of their cognitive load. Possible explanations of these findings were discussed, including methodological limitations and theoretical implications.


Included in

Psychology Commons