Implicit Attitudes Towards Individuals with Autism By College Students and the General Population

Background:  Individuals with autism spectrum disorder (ASD) demonstrate strengths and challenges. Thus, it is not surprising that some studies demonstrate that individuals with ASD are perceived negatively, whereas other research suggests positive perceptions (Swaim & Morgan, 2001). Most research has relied on self-report which is effective at evaluating explicit attitudes but is susceptible to social desirability biases. To more completely characterize the attitudes towards members of social groups, it is also important to assess implicit measures because each method predicts different types of behavior. Furthermore, measuring these attitudes in diverse samples of individuals is vital.

Objectives:  The goal of these studies was to address gaps in the literature by designing a measure of implicit bias and assessing implicit attitudes towards people with ASD. Study 1 measured implicit attitudes of college students and Study 2 focused on implicit attitudes of the general United States population. In addition, we sought to determine whether individuals who report more autistic behaviors would show less implicit bias towards persons with autism than those who report fewer autistic behaviors.

Methods:  Participants in Study 1 were 178 non-ASD college students (65 male; M = 19.15 years, 50.0% White). In Study 2, participants were 94 non-ASD individuals (50 male; M = 31.3 years, 66.0% White) who were recruited through Amazon MTurk. All participants completed a modified version of the Implicit Attitudes Test (IAT, Nosek et al., 2002) which we developed to assess attitudes towards autistic and neurotypical individuals. The Autism Quotient questionnaire (AQ; Baron-Cohen et al., 2001) measured self-reported autistic traits.

Results:  In Study 1, there was an overall implicit bias against individuals with autism (d = 0.83, SD = 0.41); this mean was significantly different from 0, t(177) = 27.09, p < .001. These findings demonstrate that participants had more positive implicit associations with neurotypical compared to autistic individuals. The IAT was significantly negatively correlated with the total AQ score, r = -.18, p = .02, such that individuals who reportedh more autistic behaviors had less implicit bias. The results of Study 2indicated implicit bias against individuals with autism (d = 0.63, SD = 0.47). This bias score was significantly different from 0, t(93) = 12.94, p < .001. As in Study 1, there was a significant negative correlation between AQ score and IAT score, r = -.45, p < .001. Thus, both college students and adults who report more autistic behaviors held less implicit bias toward those with autism.

Conclusions:  We created the first IAT to measure students’ implicit associations with neurotypical and neurodiverse individuals. In both studies, our samples generally held an implicit bias against individuals with autism. In addition, this bias was weaker for individuals who reported higher numbers of autistic behaviors. This latter finding suggests that individuals with autistic traits may be more sympathetic to individuals with autism because they have experience with and understand the behaviors displayed by those with autism.