Complicit or implicit? : the role of examiner bias in the over-identification of African American boys as emotionally disturbed
MetadataShow full item record
Despite decades of diversity focused training, African American boys are identified for special education at disproportionate rates, particularly in the disability category of Emotional Disturbance (ED). While challenges associated with poverty partially account for the over-representation, research has failed to establish a consistent relationship between poverty and special education identification. A growing volume of literature investigating the nuanced role of implicit bias in decision making has found bias to be particularly salient when criteria are ambiguous. Criteria for ED were not scientifically defined and are highly subjective. This study examined the role of implicit bias in examiner decision making when considering the eligibility of Emotional Disturbance for Caucasian and African American boys. Participants included currently practicing Licensed Specialists in School Psychology and School Psychology graduate students in Texas. Participants reviewed a mock evaluation report and made an eligibility decision in addition to rating the severity of ED. Participants then completed a survey, provided demographic information, and completed an Implicit Association Test, a measure of implicit bias. First, logistic regression and analysis of variance determined if an over-representation occurred in the collected sample by examining the rates at which boys from each race and SES category were identified as ED. Next, analysis of variance and linear regression determined if the level of examiner bias predicted the likelihood that the examiner identified the child as ED. Finally, post-hoc analyses provide additional information about the influence of pressure to provide services. Results did not support the hypotheses that African American boys and boys from a low SES family would be identified as ED at disproportionate rates. Further, examiners who demonstrated moderate to high levels of bias did not identify African American boys more frequently than Caucasian boys. Unexpectedly, examiners with moderate to high levels of bias were significantly less likely to identify ED for any student regardless of race. Post-hoc analyses found marked inconsistency in the application of ED criteria. Additionally, the examiners who reported feeling pressure from the knowledge that support services are dependent on a finding of ED were more likely to find ED and to rate the severity as higher.