Abstract
More autistic individuals are entering universities and the workforce than ever before. Yet, autistic university students experience social difficulties when transitioning to college such as social anxiety and social exclusion from their peers, and autistic adults are unemployed or underemployed. One reason for these negative experiences is the attitudes others hold toward autistic individuals. In this narrative review, we summarize the literature on explicit and implicit attitudes toward autistic individuals. Our review suggests that, while explicit attitudes toward autistic individuals as a group generally are positive, willingness to engage with autistic individuals, particularly in close proximity, may be more negative. Regarding implicit attitudes, most of which are assessed via reaction-time based methods, the findings are less nuanced. That is, regardless of the measure used, implicit attitudes toward autistic people tend to be negative in children, college students, and adults, including those who work with autistic populations. This article also considers factors that moderate these attitudes including previous contact with autistic individuals, knowledge of autism, and providing a label for autistic individuals. We review interventions designed to improve attitudes toward autistic individuals, which generally suggest that educational materials can have positive effects on explicit attitudes, but the effects of trainings on implicit attitudes are more variable. Regarding training, best practices include providing facts about autism that include the challenges and strengths associated with autism as well as providing perspectives from autistic people themselves. Finally, we provide suggestions for future directions.
Community Brief
Why is this an important issue?
Studying the attitudes that neurotypical adults hold toward autistic individuals is important, as negative attitudes can lead to negative judgments about and discrimination toward autistic people.
What was the purpose of this review?
In this review, we summarize the literature on two different types of attitudes toward autistic individuals: explicit and implicit attitudes. Explicit attitudes are conscious such that individuals are aware of and can acknowledge their attitudes. Implicit attitudes are unconscious, with individuals potentially being unaware of their attitudes.
What did the authors do to review the literature? What studies did the author find?
The authors searched online databases to find articles containing the term autism and the following words: attitude, implicit, explicit, perceptions, prejudice, stereotype, bias. The authors then summarized this research into subsections that describe the studies that have measured explicit and implicit attitudes toward autistic individuals.
In summary, what did those studies show?
Our review suggests that, while people may report that their explicit attitudes toward autistic individuals as a group are positive, their actual willingness to engage with autistic individuals may be negative. Regarding implicit attitudes, regardless of the specific test used to measure implicit attitudes toward autistic people, these attitudes were typically negative. This was the case for participants who were children, college students, and adults, including those who work with autistic populations. One other important topic we covered in this article was the identification of individual differences that might affect people’s attitudes toward autistic individuals. We found that previous studies showed that people who had more previous interactions with autistic individuals and had a greater knowledge of autism had more positive attitudes toward autistic individuals. We also reviewed interventions that researchers have used to try to improve attitudes toward autistic individuals. Our review suggests that educational materials can improve explicit attitudes but the effects of trainings on implicit attitudes are more variable. The trainings that were most effective provided facts about autism that included the challenges and strengths associated with autism and also provided personal accounts from autistic people themselves.
What are the remaining gaps in the literature?
Future research can focus on trying to identify whether interventions may be more successful for some people but less successful for others, based on individual characteristics. Studies can also test to see if the interventions have long-term effects. Furthermore, additional studies can examine whether implicit and explicit attitudes affect the way that neurotypical people interact with autistic individuals. Finally, most participants in the reviewed research studies were from Western societies so more work can be conducted in other cultures.
Based on this review, what do the authors recommend?
This paper reviews interventions that could potentially be used in business or educational settings to reduce negative attitudes toward autistic individuals. More research needs to examine whether specific components of the interventions themselves and the people participating in the interventions affect the success of the intervention.
Introduction
According to the Centers for Disease Control, current estimates are that 1 in 31 individuals is diagnosed with autism. 1 More autistic individuals are entering universities and the workforce.2–4 However, autistic university students experience difficulty in the transition to college as well as social anxiety and social exclusion from their peers.5,6 Autistic adults are more likely to be unemployed or underemployed than their neurotypical peers. 7 One factor that may affect educational and occupational success are the attitudes others hold toward autistic individuals. Autistic individuals’ perceptions of negative attitudes toward their group are associated with several negative psychological outcomes, including greater perceived isolation and poorer social relationships. 8 Given the negative impact of perceived stigma against autistic individuals, understanding the basis for attitudes toward autistic individuals—and how these attitudes can be changed—is an important research area.
In this narrative review, we use social psychological theories to frame the literature regarding attitudes toward autistic individuals. This approach incorporates defining attitudes, reviewing work identifying attitudes toward autistic individuals, considering variables that are associated with these attitudes, and examining tested and future interventions to improve attitudes toward autistic individuals. As much research in the autism field has used terms such as attitudes, perceptions, and bias interchangeably, this review clarifies these constructs and describes the different measures used to assess these variables. To this end, we also provide summaries of the findings using different measures to help elucidate seemingly contradictory research.
The review focuses on two types of attitudes: explicit and implicit attitudes toward autistic individuals. Explicit attitudes are within one’s awareness, whereas implicit attitudes are less consciously accessible and controllable.9,10 Studying both explicit and implicit attitudes is important because these two types of attitudes have been shown to predict different behaviors toward individuals who are the focus of these attitudes. That is, explicit attitudes are associated with behaviors that are more obviously directed toward an individual. Implicit attitudes are associated with more subtle behaviors (e.g., not sitting as close to somebody) toward an individual. 11 Assessing explicit and implicit attitudes toward autistic individuals is necessary to develop a more complete understanding of the attitudes toward these individuals, to understand the range of behaviors that are experienced by autistic individuals, and to develop interventions that remediate the different types of negative attitudes that exist against autistic individuals.
To obtain articles for our narrative review, we searched PsycINFO for articles containing the term autism and the following words: attitude, implicit, explicit, perceptions, prejudice, stereotype, bias. There were no date restrictions. We conducted backward and forward searches with relevant empirical journal articles published in peer-reviewed journals that used human subjects research to investigate implicit and explicit attitudes, perceptions, and judgments about autistic individuals until we reached saturation. The subsections that follow describe the evidence regarding explicit and implicit attitudes toward autistic individuals along with variables related to these attitudes. We also include a section reviewing research attempting to improve implicit and explicit attitudes toward autistic people. Finally, we discuss gaps in the literature and provide suggestions for future directions.
Explicit Attitudes Toward Autistic Individuals
A range of methods have been used to assess explicit attitudes toward autistic individuals, most of which measure cognitive representations such as stereotypes or judgments associated with autistic people. These measures have used open-ended responses,12,13 Likert-type scales,14,15 and judgments about autistic individuals described in vignettes or viewed in video clips.16,17 Other studies examine behavioral intentions, 18 which represent participants’ comfort with and intentions to interact with autistic people in the future. Together, this research has examined explicit attitudes toward autistic children, 19 university students, 13 and non-student adults. 20 Participants include neurotypical and autistic children, 21 university students, 22 non-student adults, 23 and teachers. 24 In this section, we review research that used different methods to explore explicit attitudes toward autistic people.
One approach to begin to understand the characteristics that people associate with autistic people is to ask participants open-ended questions regarding what behavior or traits they feel are characteristic of autistic individuals. Using this method, qualitative work with university student participants demonstrated that commonly identified categories tend to be related to traits associated with poor social abilities, being dependent on others, being introverted, intelligence, poor communication, and awkwardness.12,13,25 These responses suggest that university students are aware of the differences in social skills that autistic people may display relative to neurotypical individuals and also are aware of autism as a spectrum disorder.
In addition to examining open-ended responses, researchers also assessed explicit attitudes toward autistic people using Likert-type scales in which participants indicate the degree they think a trait is associated with autistic people in general, using positive and negative statements about autistic people. One such measure that was created to assess teachers’ feelings about autistic children, the Autism Attitude Scale for Teachers, 15 consists of seven 5-point Likert-type items with items such as “Normal children and autistic children should be taught in separate schools.” Using this measure, pre-service teachers and teachers in the United States indicated relatively positive explicit attitudes toward autistic children, although this scale focused not only on teachers’ feelings about autistic students but also their knowledge of autism.19,24 Similarly, the Attitudes to Autism scale 14 contains 10 statements asking participants’ agreement or disagreement to questions related to positive (e.g., creative) and negative (e.g., unsociable) traits; one study using this measure demonstrated that university students had relatively positive views of autistic children. 22 The same sample of participants also showed generally positive views of autistic individuals (instructions did not specify children or adults) on a 7-point Likert-type scale assessing positive (e.g., good, happy) and negative (e.g., bad, difficult) attributes. 22
Although studies using positive and negative traits have revealed relatively positive attitudes, as measured by cognitive associations, toward autistic children and autistic people, there is evidence that autistic people are viewed as less human than neurotypical people. One such study asked adolescents and young adults to rate on a 7-point Likert-type scale the degree to which personality traits were associated with autistic and non-autistic people. 20 Results indicated that ratings of autistic and non-autistic people did not differ on human nature traits (e.g., friendly, impatient), but non-autistic people were rated higher on human uniqueness traits (e.g., broad-minded, ignorant). Similarly, another study demonstrated that adults categorized autistic individuals as more animal-like and child-like than non-autistic individuals. 26 The authors argue that these findings are evidence that people tend to hold dehumanizing attitudes about autistic individuals.
Further evidence demonstrating that autistic people are viewed more negatively than neurotypical people comes from research assessing perceptions of autistic individuals rather than autistic people as a group. This research typically features autistic individuals displayed in videos or described in vignettes. Several studies used short 10-second video clips depicting autistic and neurotypical individuals followed by judgment ratings. Results indicated that both neurotypical and autistic participants viewed autistic individuals more negatively on the First Impressions Scale, 17 indicating that they were more awkward, less attractive, and less dominant than the non-autistic individuals portrayed,16,17,27,28 although diagnosis disclosure and individual characteristics of the participants affected first impressions (see the Correlates section). This method was also used to assess the perceptions of autistic children. Boucher and colleagues (2023) had university students evaluate autistic and non-autistic children on positive and negative characteristics using the First Impressions Scale after viewing 9–14-second video clips of them talking about their favorite things. The autistic children were rated significantly more negatively than the non-autistic children on awkwardness, trustworthiness, aggression, and likeability. In another study, adult participants rated an autistic six-year-old boy more negatively (e.g., more aggressive, less nice) when he displayed stereotypical autistic behaviors (i.e., moaning, temper tantrum) than when he displayed neutral behaviors (i.e., using a computer, talking). 29 Similarly, third- and sixth-grade children reported less positive attitudes as assessed by an adjective checklist (e.g., smart, neat, dumb, sloppy) to the same child engaging in hand flapping, rocking, and echolalia compared with when the child displayed none of these behaviors. 21 Autistic children were also rated as less desirable friends based on 50-second video clips of children talking about themselves but with no audio. 30 Even when videos were as short as 1 second, neurotypical adults rated autistic children as more socially awkward than non-autistic children, suggesting that negative attitudes toward autistic individuals may at least partially result from nonverbal cues. 27 To improve ecological validity over using videos of autistic individuals, Lipson et al. (2020) examined perceptions of a trained actor who either displayed behaviors consistent with autism or without autistic behaviors in a face-to-face interaction. They found that neurotypical university students had more negative perceptions of the confederate (e.g., “I would avoid this person”) when he portrayed stereotypically autistic compared with no autistic behaviors. In fact, the more autistic behaviors participants perceived the confederate to have, the more negatively they rated the confederate’s social skills. 13 Together, these results suggest that while attitudes toward autistic children and adults may be positive when the entire group is evaluated, attitudes toward individual autistic children and adults as seen in videos can be negative, even at short exposures.
In addition to examining the perceptions of individuals depicted in videos, in vignettes, and in person, researchers have also examined participants’ desire for future interactions with autistic individuals in varying levels of intimacy. Studies examining short video clips of adults found that participants indicated that they would be less likely to hang out with, sit next to, and start a conversation with the autistic individuals portrayed in the video clips than the non-autistic individuals.17,27,28 Studies examining attitudes toward children revealed that university student participants were also less willing to interact with and had lower comfort levels with autistic children 31 and college students. 32 Importantly, for attitudes toward college students, while scores on each item ranged from the lowest to the highest score on the 5-point Likert-type scale, the mean of participants’ attitudes was at the midpoint, showing moderate openness to their autistic peers. 32 A follow-up study with the same population found that university students’ attitudes toward their autistic peers as measured by openness were more positive after a 5-year period, 33 suggesting that attitudes may change over time or may be different between cohorts. In a study assessing the degree of closeness in future interactions with autistic individuals, while more than 60% of participants would feel “okay” or “very okay” having an autistic person as a neighbor, colleague, or friend, less than 40% of participants would be “okay” with having an autistic person as a boss, and less than 20% would be okay having a romantic relationship with an autistic individual. 23 These results suggest that, while overall attitudes toward autistic individuals as a group may be positive, willingness to engage with autistic individuals, particularly in close proximity, may be more negative.
Implicit Attitudes Toward Autistic Individuals
Several different measures have been used to assess implicit attitudes toward autistic individuals. The most commonly used measure of implicit attitudes in social psychology is the Implicit Association Test (IAT). 10 The IAT was designed to examine the automatic cognitive associations that operate at least partially outside of conscious awareness. The task presents participants with picture or word stimuli that can represent different social groups (e.g., autistic people, neurotypical people) and valence (e.g., positive, negative). Participants press one of two keys on a keyboard to sort the stimuli into their corresponding category (i.e., group, valence). First, participants complete trials in which they sort positive and negative words into the valence categories and other trials in which they sort stimuli related to the category (e.g., autistic people, neurotypical people) into the corresponding social categories. Next, the two categories are combined (e.g., positive + autistic; negative + neurotypical) and participants sort all stimulus words into the combined groups. Following this, the categories for positive and negative words switch and participants end with critical trials in which the two categories are again combined but with the opposite pairing (e.g., negative + autistic; positive + neurotypical). Faster times to categorize stimuli are thought to reflect closer cognitive associations; the differences in associations between categories are thought to assess participants’ degree of positivity toward one group compared with another group. 10 Although there are well-documented concerns with reliability and predictive validity that are beyond the scope of this review, 34 there is evidence that the IAT can reliably produce large effects and predict relevant behavioral outcomes. 35
Dickter and colleagues (2020) adapted the original IAT to examine implicit attitudes toward autistic compared with non-autistic individuals. Their categorical words for the autistic category were autistic, spectrum, ASD, and Asperger’s; for the non-autistic category, the words were typical, nonautistic, usual, non-ASD. The words for the positive and negative categories were the words used in most social psychological IATs (e.g., marvelous, superb, horrible, awful). 10 Studies using this version of the autism IAT reported that participants showed more negative associations with the autistic category than the non-autistic category, as demonstrated with university samples13,36 and non-student adults.37,38 Lacruz-Pérez and colleagues (2023) translated this IAT into Spanish and also found evidence of negative autism associations in a sample of 50 pre-service teachers in Spain. In a single-category version of the IAT, which does not use two social categories at the same time, Kim et al. (2023) had participants categorize stimulus words into three categories: autism (“dislikes unexpected changes,” “repeats body movements”), good (e.g., “joy”), and bad (e.g., “angry”). Consistent with traditional versions of the IAT, results indicated that the Korean and American adult participants showed negative implicit associations with autism. 39 Together, these studies demonstrate that samples of adults from multiple cultures show negative implicit associations toward autistic individuals using IATs with word stimuli.
Yokota and Tanaka (2024) created an IAT with picture stimuli rather than word stimuli. In this task, autistic targets are portrayed in pictograms exhibiting behaviors related to social exclusion, flapping arms, and attention to detail. Participants classify the pictures into healthy versus disorder and good versus bad categories. Consistent with previous research using words, university students showed negative implicit associations with autistic targets.40,41 Interestingly, the neural activation pattern of the right amygdala to the pictures of autistic compared with the non-autistic pictograms was correlated with implicit associations as assessed with the IAT. 41
Another task that has been used to assess implicit attitudes toward autistic people is the Implicit Relational Assessment Procedure (IRAP). Proponents of the IRAP suggest that while the IAT measures only associations between concepts, the IRAP allows for a focus on conceptual attitudinal relations independently. That is, while the IAT presents autistic and non-autistic categories at the same time, the IRAP presents each one individually on different trials. Kelly and Barnes-Holmes (2013) adapted this task by having participants complete six test blocks with 24 trials, in which 12 target stimuli showing six positive words that they deemed were consistent with perceptions of “normally developing” individuals (e.g., happy, easy) and six negative words that they chose to be consistent with perceptions of autism and used the answer choices “similar” and “opposite.” In the IRAP, participants categorize the target stimuli as similar or opposite based on the categories “Autistic Spectrum Disorder” and “Normally Developing”; as with the IAT, blocks with associations of autism and negative are counterbalanced with blocks with autism and positive. Participants who respond more quickly to autism-negative-similar trials compared with autism-negative-opposite trials are considered to have negative implicit associations with autistic people. Using this method, a number of studies demonstrated that professionals working with autistic populations, applied behavior analysis tutors, K-12 teachers in mainstream classrooms, and undergraduate students showed greater positivity toward normally developing than autistic children.14,22,42
More recently, Aubé and colleagues (2021) used an approach-avoidance task, the Visual-Approach/Avoidance-by-the-Self-Task (VAAST) to measure implicit attitudes. They argued that, in addition to measuring spontaneous reactions to stimuli with high reliability, the VAAST is also easier for certain populations such as children to use. The VAAST requires participants to press a button or move a joystick forward or backward to make the stimulus larger or smaller. In the autism VAAST utilized by Aubé et al. (2021), child participants first watched a 6-minute video introducing them to three autistic children with intellectual disabilities displaying behaviors consistent with autism and three neurotypical children identified as children with “a mental disability” or “without a mental disability.” During the VAAST, participants were shown pictures of the autistic and neurotypical children from the videos and had to push a key to move forward (i.e., approach) or backward (i.e., avoid) pictures of children with and without a mental disability. The results indicated that participants had more negative implicit attitudes toward the autistic children than the non-autistic children. Interestingly, while they showed that older children had more positive explicit attitudes as measured by perceived similarity and behavioral intentions toward the autistic children, implicit attitudes did not differ as a function of age. 18
Taken together, these studies reveal that implicit attitudes toward autistic people, as measured by the strength of automatic associations assessed by a variety of reaction-time-based tasks, tend to be negative in children, college students, and adults, including those who work with autistic populations. These findings have important implications for the experiences of autistic people, as implicit attitudes can predict subtle forms of behavior, including judgments, engagement, and exclusion. 11
Correlates
Research has identified a number of variables that are linked to biases toward autistic individuals. One construct that has been extensively studied is previous contact with autistic individuals. Personal contact is typically measured using self-report items that gauge the amount of time participants have spent with autistic individuals. Research has demonstrated that university students with greater previous contact have more positive explicit attitudes toward autistic individuals. 43 Furthermore, university students with a first-degree autistic relative report greater openness to autistic peers than those without an autistic relative.32,44 Importantly, in addition to associations with attitudes and openness, greater previous contact has also been associated with more positive behavioral intentions toward autistic individuals. For example, greater previous contact was associated with higher likelihood of future desires to socially interact with an autistic individual 23 and with higher likelihood of signing up to volunteer to work with an autistic individual. 44 Among school-age children, greater previous contact has been associated with lower tolerance for bullying toward autistic individuals. 45 In summary, increased self-reported previous contact with autistic individuals is associated with more positive attitudes and behavioral intentions toward autistic individuals. However, additional research is needed to investigate the circumstances under which contact leads to positive attitudes toward autistic individuals as opposed to less negative attitudes toward autistic individuals.
Knowledge of autism is another variable that has been investigated in relation to explicit and implicit attitudes. Sasson and Morrison (2019) found that neurotypical university students with greater knowledge about autism, as assessed by the Autism Knowledge Scale, 46 had more positive attitudes toward autistic individuals. These findings were further confirmed by a meta-analysis showing that knowledge of autism is associated with attitudes toward autistic individuals. 39 These results are consistent with ideas that characteristics of the participant raters, such as knowledge of autism and previous contact with autistic individuals, may be more important in attitudes and behavioral intentions toward autistic individuals than the behaviors of autistic individuals. 47 This conclusion could lead to the idea that the autistic traits of the rater may be an important factor in attitudes and behavioral intentions toward autistic individuals, as those raters with higher autistic traits may have greater understanding of autistic behaviors. However, the literature is mixed on this point, with evidence showing that greater self-reported autistic behaviors as measured by the Autism-Spectrum Quotient (Baron-Cohen et al., 2001) were associated with less implicit bias toward autistic individuals, 37 but other studies show that autistic individuals have more negative attitudes toward autistic compared with neurotypical individuals portrayed in video clips. 16 Clearly, the relationship between a rater’s autistic characteristics and the attitudes that raters hold toward autistic individuals needs further investigation to understand the underlying mechanisms of attitude change and to understand when attitudes shift to being positive toward autistic individuals.
A growing body of research has examined whether labeling a person as autistic would affect attitudes toward the autistic individual. These studies generally demonstrate that labels of autism improve perceptions of an individual displaying behaviors or characteristics consistent with autism described in a vignette compared with when the person is not labeled as autistic28,29,48,49 as well as willingness to interact with a hypothetical coworker. 50 Similarly, neurotypical participants rate autistic individuals portrayed in videos more positively when they are labeled as autistic compared with when no label is provided.16,28 However, other studies using vignettes, videos, or in-person interactions conveying autistic behavior demonstrated that attitudes were not affected by diagnosis disclosure.13,21,33,51 Finally, research has demonstrated that providing descriptive and explanatory information about autistic behaviors improves attitudes and behavioral intentions toward autistic children in neurotypical child and adult participants.52–56 The mechanism by which labeling is associated with more positive attitudes toward autistic individuals is not well-understood. The label may cause individuals to self-present in a more positive manner in the moment, but may not lead to enduring attitude change. One strategy to begin to address this concern would be to conduct longitudinal studies to assess whether changes in attitudes are long-lasting.
Demographic differences have also been examined, with nuanced results. For adults rating a video of a 6-year-old autistic boy, men were more likely to evaluate the child severely compared with women, although this gender difference was no longer present when explanatory information about autism was provided. 29 A study with middle school students showed that female students had a more positive attitude toward a video of an autistic peer compared with male students, regardless of the type of written message that accompanied the video. 52 However, another study using third- and sixth-grade children found that females but not males rated a video of an autistic child more negatively compared with a control condition when the child did not display autistic behaviors. 21 Using a sample from Denmark, one study reported that male adult participants identified a lower percentage of autistic traits compared with females when given traits based on the International Classification of Diseases and were less willing to interact with autistic people. 23 These findings are in contrast with studies with American college students that demonstrated that males had more positive attitudes about and higher openness to interacting with autistic characters portrayed in vignettes.32,49 Thus, the evidence is mixed regarding any gender differences in attitudes toward autistic individuals.
Other demographic variables, specifically age and educational attainment, do not seem to be linked to attitudes toward autistic individuals. 23 The role of culture in developing attitudes toward autistic individuals will be an important variable in future studies. For example, recent studies showed greater explicit and implicit bias in South Korean participants compared with American participants, which the authors suggest may be related to a more assimilationist South Korean culture.26,57
Strategies to Reduce Biases Toward Autistic Individuals
Given the relatively consistent findings that individuals hold negative explicit and implicit attitudes toward autistic people, with exceptions dependent on the task and sample, along with the established potential negative repercussions for autistic individuals, several research groups have investigated strategies to reduce these biases. Two main theoretical frameworks have been used to inform interventions aimed to reduce bias toward autistic individuals. First, researchers have viewed attitude change techniques as a process of persuasive communication that provides descriptive information about autistic individuals, knowledge-based information on autism, and guidance on how to effectively interact with autistic people. Second, Allport’s (1954) Contact Theory has been used as a framework whereby positive attitude change is accomplished by positive intergroup interactions. 58 We review research using each of these perspectives below.
Knowledge-based interventions
As discussed above, with knowledge of autism being consistently associated with attitudes toward autistic individuals, several strategies to improve attitudes have focused on educating participants about autism. A first attempt to change neurotypical third–sixth graders’ explicit attitudes toward autistic children by presenting them with information about autism after viewing a video of a child displaying autistic behaviors did not affect attitudes. 21 Subsequent studies providing additional information about autism rather than simple facts about behaviors related to autism, however, were more successful. In a study presenting videos to neurotypical fourth–sixth graders, although attitudes toward autistic children were unaffected by the videos, behavioral intentions toward autistic children were more positive when the videos included peer strategies (i.e., ways in which children could interact appropriately with individuals displaying “social challenges”) compared with when the videos only included information about the strengths that autistic children may possess (e.g., perceptual abilities). 56 Pre-service teachers’ explicit attitudes toward autistic students, as assessed with the Autism Attitude Scale for Teachers, 15 were improved by a short-term training about autism that included information about the diagnostic criteria of autism, guidelines for intervention with autistic students, and awareness about including autistic children in mainstream classes, while implicit attitudes measured with a single-target IAT were unaffected. 59 With high school students, a 50-minute educational presentation focused on differences between autistic and non-autistic individuals related to communication, sensory processing, behavior, interests, and empathy produced a modest improvement in explicit attitudes toward autistic adults portrayed in videos, including finding them more likable and attractive, compared with individuals who had not viewed the presentation. In a similar but longer training, seventh–ninth-grade girls exposed to eight 50-minute sessions focusing on educating students about the impairments and presentation of higher functioning autism had more positive knowledge, attitudes, and behavioral intentions toward autistic peers. 60 Together, these studies demonstrate that short- and long-term interventions that provide educational information detailing the specific traits of autistic individuals can improve attitudes and behavioral intentions in children and adolescents.
Several approaches to improve the attitudes of non-autistic adults toward autistic adults have been examined. Gillespie-Lynch and colleagues (2022) developed two trainings that educated participants about autism, characteristics and misconceptions of autism, disparities in access to care, and factors that contribute to autism while providing strategies to support autistic students. Results indicated that both trainings, a participatory one providing videos of autistic students who developed the training sharing their experiences and a non-participatory training providing videos that were more information-based (e.g., Ted talk), improved IAT scores and knowledge about autism. 61 In a subsequent study, the participatory intervention improved autism knowledge, explicit stigma, and attitudes toward inclusion. Another autism acceptance training, presented in a 38-slide PowerPoint presentation, reviewed the characteristics of autistic people, including challenges and strengths, using education and first-person accounts, improved aspects of explicit attitudes (e.g., more positive impressions, less stigma), but not implicit attitudes toward autistic adults compared with control conditions. 62 An assessment of the effectiveness of a website developed by Sesame Street that provided general information about autism with videos of autistic children to parents with and without autistic children revealed that parents without autistic children and those with negative explicit attitudes toward autistic individuals showed a reduction in implicit bias toward autistic children after viewing the site. 38 One study with non-autistic undergraduate students found more positive explicit and implicit attitudes using the IRAP following brief exposure to an informational video and text about autism 22 compared with control conditions.
In summary, the available literature suggests that educational materials regarding autism typically have positive effects on explicit attitudes but more inconsistent effects on implicit attitudes, which may reflect differences in methodological procedures for assessing implicit attitudes or the less malleable properties of implicit attitudes. Best practices seem to include providing not only basic information about autism and the challenges and strengths associated with autism but also the inclusion of perspectives from autistic people themselves. Importantly, recent interventions have taken care to include autistic individuals in the development of training materials. While there is encouraging research on the use of educational strategies to improve attitudes toward autistic individuals, 63 some recent reviews articulate some important limitations of the literature that need to be addressed in future research. 64 These limitations include a narrow focus regarding the age, race, and gender of the target used to give participants exposure to an autistic individual. Additional identified issues included the absence of measures of knowledge of autism and of measures of observable behaviors rather than relying exclusively on self-report measures. Another systematic review identified that intervention effectiveness was sometimes not described accurately and limitations of study design were not articulated. 65 These authors mention that internal consistencies of measures of outcome variables, risk of participant social desirability bias in anti-stigma programs, and inclusion of autistic individuals in developing interventions should all be important considerations for future research.
Contact-based interventions
As reviewed above, individuals with greater previous contact with autistic people tend to have more positive attitudes. 43 This finding is consistent with Allport’s (1954) Contact Theory in which intergroup contact is thought to reduce prejudice. Research examining improvements in explicit and implicit attitudes toward members of different social groups has found that a reduction on the dependence on stereotypes, reduced anxiety, enhanced knowledge, and increased empathy and perspective taking are mediators responsible for this attitude change. 58 As such, interventions aiming to improve attitudes toward outgroup members have demonstrated that induced positive intergroup contact can be effective. 58 Researchers have determined that imagining a positive intergroup interaction can also be effective at improving attitudes toward outgroup members, 66 which is important, as interpersonal contact may not always be possible or positive. When adapted to an imagined interaction with an autistic individual, explicit 43 and implicit 36 attitudes toward autistic individuals were more positive following an imagined contact scenario with a peer labeled as autistic displaying behaviors associated with autism. Moreover, following an imagined contact scenario, explicit attitudes were more positive when the name of the individual was a stereotypical male compared with a stereotypical female; intergroup anxiety mediated this relationship. 67 Another recent line of research related to exposure to autistic individuals has begun to demonstrate that, under some conditions, portrayals of autistic individuals in popular media, including television shows and books, can improve attitudes toward autistic individuals and that this improvement is mediated by liking for the autistic character.68–70 An area for future investigation is whether some of the same mechanisms underlying the effects of imagined contact on attitudes also apply to changes in attitudes toward autistic individuals following portrayals of autistic individuals in popular media. In summary, interventions that involve imagining positive interactions with autistic individuals or that involve portrayals of autistic individuals appear to have beneficial effects on attitudes toward autistic individuals.
Future Directions
The present review summarizes key findings regarding attitudes toward autistic individuals. One challenge with the field is that terms such as explicit and implicit attitudes are umbrella terms made up of numerous constructs, including perceptions, judgments, associations, and bias. Researchers should take care to ensure that they accurately define the construct of interest as well as the measurement technique that operationally defines the variable of interest.
As our section on strategies to reduce bias toward autistic individuals describes, some trainings have demonstrated efficacy in reducing negative attitudes in children,56,60 teachers, 59 and college students.22,62,61 Future research should examine moderators of the effect of these trainings on improving attitudes. This is particularly important as some studies show that explicit attitudes were changed while implicit attitudes were not, although others found positive changes in implicit attitudes. For example, trainings may be more effective for people concerned about their personal tendency to hold prejudice against autistic individuals or who have less previous contact with and more negative attitudes about autistic individuals. In addition, whether participants subscribe to more of a medical model or social model of disability may impact whether trainings affect participant attitudes toward autistic individuals. Furthermore, as mentioned above, gender may moderate the effectiveness of trainings; other demographic variables should also be examined. Indeed, trainings that have successfully changed implicit attitudes toward minoritized groups (e.g., racial groups) have focused on “breaking the habit” of implicit bias through replacing stereotypical responses with non-stereotypical responses, thinking about counter-stereotypical exemplars of the minoritized group, focusing on individual characteristics of individuals from the minoritized group, and taking the perspective of a member of the minoritized group. 71 This training is most effective at reducing implicit bias in individuals who were most concerned about discrimination. In addition, given that contact reduces bias toward autistic individuals, trainings that involve first-person accounts of autistic people may be especially effective. During development, future interventions should be theory-driven and informed by social movements such as the neurodiversity perspective. Importantly, longitudinal studies are needed to understand the duration of attitude changes following an intervention.
Given that autistic individuals experience social exclusion from their neurotypical peers, 72 it is important to examine how explicit and implicit attitudes are related to the way neurotypical individuals interact with autistic individuals. As implicit and explicit measures are thought to be associated with different behaviors toward autistic individuals, examining whether attitudinal measures predict behavior toward actual autistic people is imperative. For example, social psychological research has demonstrated that implicit racial attitudes predict nonverbal behavior in intergroup interactions, while explicit racial attitudes predict verbal behavior. 11 One study applying this research to autism revealed that college students with more implicit bias as measured by the IAT showed more negative nonverbal behavior (i.e., leaning away from) in a face-to-face interaction from a peer identified to be autistic who displayed behaviors stereotypical of autistic people. 13 Other research demonstrated that hiring agents’ negative explicit attitudes toward autistic individuals predicted their selection of a qualified autistic individual for employment. 73
Another future direction includes characterizing attitudes toward autistic individuals with different characteristics. For example, using an IAT, one recent study found that implicit associations were stronger between male names and autistic characteristics compared with female names and autistic characteristics. 74 Future research can also examine attitudes toward autistic males and females separately, as there is evidence that there can be differences in attitudes based on gender.
Future research may also aim to characterize implicit bias at the geographic level to help understand the context-based systemic biases against autistic individuals. For example, social psychological research has demonstrated that more implicit racial bias was seen in counties that were more dependent on slavery before the Civil War 75 and that less racial implicit bias was seen in areas with greater racial diversity. 76 Researchers may aim to understand attitudes toward autistic individuals through examining systemic historical or demographic factors.
In addition, although explicit and implicit attitudes are thought to be separate constructs, it is important to examine the relation between them to thoroughly understand attitudes toward autistic individuals. Many of the studies reviewed above using both implicit and explicit measures did not report relations between them. Some research has demonstrated that there are small but nonsignificant correlations between an explicit measure, the Societal Attitudes Towards Autism Scale, 77 and the Autism IAT in studies with adults (rs from 0.07 to 0.20), in that participants with more negative implicit attitudes showed more negative explicit attitudes.13,37 Similarly, research revealed small and nonsignificant correlations between a single-target IAT and the Social Distance Scale in American and Korean adults 57 and the Autism Attitude Scale for Teachers in 50 pre-service teachers in Spain. 59 While IAT research has largely revealed small to no relations between the IAT and various explicit measures, other research has demonstrated significant correlations between the IRAP and explicit measures in child participants 14 and college students (rs from 0.29 to 0.35). 22 The fact that the relations between the explicit and implicit attitude variables depend on the specific measures used speaks to the importance of continuing to examine these relations using a variety of different methods to fully understand the nature of the relations between these variables. Finally, the majority of the work reviewed above has been conducted with Western young adults (many of whom are college students) and children as participants. Future research should take care to recruit participants who are more diverse in age, culture, socioeconomic status, and education level. In addition, research needs to expand regarding the characteristics of the autistic individuals being assessed. For example, many studies use White males 13 or names stereotypically associated with White males as the autistic individual used to assess attitudes. Research needs to expand by assessing attitudes toward female and non-White autistic individuals. 67 In addition, given the importance of intersectionality as it applies to attitudes toward group members, examining attitudes toward autistic individuals from different socially constructed identities such as race, gender, and sexual orientation is necessary.
Footnotes
Author Disclosure Statement
No conflicts of interest to declare.
Funding Information
No funding was received for this article.
Authorship Confirmation Statement
C.L.D.: Conceptualization (equal), writing—original draft (equal), and writing—review and editing (equal). J.A.B.: Conceptualization (equal), writing—original draft (equal), and writing—review and editing (equal). This article has been given solely to this journal and is not published, in press, or submitted elsewhere.
