Abstract
The current study sought to characterize implicit bias toward children with autism and examine whether viewing educational materials about autism would change attitudes toward children with autism. A website developed by Sesame Street containing information about autism and resources for families was distributed to parents of children with autism (n = 473) and parents of children without autism (n = 707). Pre- and post-test measures of implicit bias toward children with autism; explicit attitudes and knowledge about autism; and parenting confidence, strain, and stigma were completed before and after the website was presented. Results indicated that parents of children with autism showed less implicit bias compared with those of non-autistic children during the pre-test, but the groups did not differ at the post-test. Parents without autistic children and those with more negative explicit attitudes showed a greater reduction in implicit bias from the pre- to the post-test. In addition, for parents of children with autism, a more positive change in explicit attitudes and increased knowledge from the pre- to the post-test was associated with more empowerment at the post-test. Together, our findings suggest that the online educational resources can reduce implicit bias against children with autism and help mitigate some of the psychological issues associated with parenting children with autism.
From 2000 until present, the prevalence of autism spectrum disorder (ASD) per 1000 children has more than doubled from 6.8 to approximately 14.6 (Centers for Disease Control and Prevention, 2018). Given this increase in prevalence, research regarding attitudes toward this population is critical to understand ways in which children with autism may be disadvantaged. Indeed, the social exclusion, humiliation, and neglect that individuals with ASD often face (Attwood, 1998; Gray, 1993; Peyton, 2004) may contribute to the higher rate of affective disorders in samples of individuals with ASD compared with non-ASD samples (Tantam, 2000), including depression (Hedley & Young, 2006) and anxiety (S. W. White et al., 2011). Therefore, interventions aimed at increasing understanding of autism and reducing negative attitudes toward children with ASD are important. The current study aimed to assess both implicit and explicit attitudes toward children with ASD in parents with and without children with autism. Furthermore, it aimed to assess an online resource created to provide information about autism and to study the effects of this resource on parents with and without children with ASD.
Explicit attitudes toward children with autism
Previous research examining attitudes toward children with ASD has focused on perceptions of children’s behaviors. For example, studies show that children with ASD displaying autistic behaviors are generally judged more negatively by their peers (Campbell et al., 2004; Swaim & Morgan, 2001) and by young adults (Iobst et al., 2009) than those showing behaviors not associated with autism. Experiencing negative judgments by others may account for findings that 95% of parents report that their children with ASD are stigmatized and experience at least some rejection by their peers (Kinnear et al., 2016). However, research has demonstrated that individuals with more knowledge of and previous experience with individuals with ASD have more positive attitudes toward children with autism (e.g. Mahoney, 2008; Nevill & White, 2011; Park et al., 2010), suggesting a potential way to reduce negative attitudes toward children with ASD.
Implicit attitudes toward children with autism
The studies described above relied exclusively on self-report measures. Although self-report methods are effective at measuring explicit attitudes that people readily identify, such measures can suffer from social desirability biases, particularly in situations in which participants are sensitive to social norms regarding stereotypes about social groups (e.g. Dovidio & Gaertner, 2000). Indeed, self-report measures of attitudes toward individuals with disabilities are susceptible to socially desirable responding (Hergenrather & Rhodes, 2007; Pruett & Chan, 2006). In addition, implicit attitudes can be activated outside of awareness so that individuals do not recognize their bias (e.g. Devine & Monteith, 1999; Payne, 2001). Implicit attitudes predict subtle forms of discrimination, such as behavioral rejection or discomfort (e.g. Dovidio et al., 2002; Fazio et al., 2005). Thus, identifying implicit attitudes toward autistic children can help elucidate the mechanisms that may lead to discriminatory behavior toward these children.
Implicit attitudes toward individuals of a given group are often measured by the Implicit Association Test (IAT; Greenwald et al., 1998). The IAT has been shown to be a valid and reliable measure that predicts behavior in numerous domains (Greenwald et al., 2009), although effect size estimates of predictive validity differ between meta-analysis strategy (Greenwald et al., 2015; Oswald et al., 2013). Disabilities-focused IATs have generally found that people show an implicit bias against words and phrases related to mental illness (Pruett & Chan, 2006; Stier & Hinshaw, 2007; Thomas et al., 2007; M. J. White et al., 2006). One set of studies developing an autism IAT demonstrated that college students and non-college adults, particularly those who were low in the degree of autistic traits they self-reported, showed an implicit bias toward their peers with autism (Dickter et al., 2020). Previous research has not examined adults’ implicit attitudes toward children with autism, but given these previous results, it is likely that adults hold negative implicit bias toward children with autism.
Psychological outcomes for parents with children with autism
Although examining attitudes toward children with autism in all adults is important, examining these attitudes in parents with children with ASD is of particular interest due to the psychological barriers that parents of children with autism may face. Research has shown that having a child with ASD can lead to isolation, strain, and mental health problems in parents (Baker-Ericzén et al., 2005; Farrugia, 2009; Gray, 1993; Kinnear et al., 2016; Woodgate et al., 2008); finding ways to improve these outcomes is crucial. The Family Adjustment and Adaptation Response (FAAR) model (Patterson, 1988) provides a theoretical framework to examine how parents of children with autism can use resources and coping behaviors to deal with the isolation, stigma, and strain that can be associated with parenting children with autism. Research applying this model to parents of children with disabilities suggests that gaining access to resources can lead to greater feelings of empowerment (i.e., greater feelings of parental competence), which in turn can increase positive outcomes in parents of children with disabilities (Nachshen & Minnes, 2005; Scheel & Rieckmann, 1998; Weiss et al., 2015). Empowerment is associated with higher use of services, more awareness of social support, and higher self-efficacy (Wakimizu et al., 2011). Thus, a resource that improves attitudes toward children with ASD and increases knowledge about autism may enhance parental empowerment and mitigate some of the psychological issues associated with parenting children with ASD. One critical factor is access to resources and services, which is dependent on many factors, including income and siblings requiring care (Karp et al., 2018). A resource that is free and easily available over the Internet could provide relatively accessible information to parents of children with ASD.
The current study
The current study recruited groups of parents with and without a child with autism to achieve three goals. First, it sought to characterize implicit bias toward children with autism. Consistent with previous research with college students and the general population (Dickter et al., 2020), we expected that participants would have a negative implicit bias toward children with ASD. However, because parents of children with autism report better understanding of children with disabilities and of autism in general (Pakenham et al., 2004), we expected this bias to be more negative in parents without children with ASD.
Second, this study sought to examine whether viewing educational materials about autism would change parents’ implicit bias toward young children with autism. To promote knowledge and acceptance of children with ASD age 2 through 5 years, Sesame Workshop created an online initiative, Sesame Street and Autism: See Amazing in All Children, henceforth referred to as See Amazing. The free website includes general information about autism and links to resources for families and the public. The website and the resource materials were developed with the goals of reducing stigma, increasing knowledge about ASD, and building positive perceptions about ASD. All materials were created with input from autistic self-advocates, people who serve the autism community, and parents. To examine the impact of the See Amazing resources, we measured implicit bias before and after parents viewed the resources, expecting it to be less negative after interacting with the See Amazing material. In addition, we examined whether implicit bias change was affected by the degree of implicit and explicit bias participants held at the beginning of the study as well as whether there were different shifts for parents of children with autism versus parents without children with autism.
Finally, in addition to reducing implicit bias, we sought to test whether potential changes in parental competence were related to changes in implicit bias, acceptance, knowledge, familiarity, strain, and stigma in parents with children with ASD. We hypothesized that a positive change in attitudes toward autism, increased knowledge about autism, increased acceptance, and reduced strain and stigma as a result of the website would increase feelings of empowerment in parents of children with ASD.
Method
Participants
This study was conducted as part of a larger evaluation of See Amazing initiative (Anthony et al., 2020). Eligible participants included adult parents or primary caregivers of a child under 6 years of age. Potential participants were sent a link to an evaluation survey and asked to complete an initial set of screening questions regarding their child’s age and whether their child was autistic. If their responses were appropriate for either the group of parents with a child with ASD or the group of parents without a child with ASD, participants were directed to an informed consent form.
Parents of children with ASD
Parents who had a child with ASD were approached across the nation through a variety of sources: emails targeting parents whose children had been seen at ASD clinics and had indicated a willingness to participate in research, autism-related websites, flyers in clinics and service programs, autism-relevant listservs, and Facebook posts. A link to the baseline survey was accessed by 903 parents of whom 312 met eligibility criteria, consented to being in the study, and completed 90% of the baseline survey, a requirement for continuing in the study. Participants lived in 311 different zip codes. Approximately, half of the participants had either heard of the See Amazing website or had some familiarity with it, but only 4% had viewed the entire site. As depicted in Table 1, over 80% of participants were female and about three quarters of their children were male with a mean age of 4.5 years. The sample was 73.9% White and 9.1% Black, which is about 10% higher and 5% lower, respectively, than the national census rates. The percentage of Hispanic/Latino participants (19.0%) was close to national estimates. Participants represented broad educational attainments—about 50% reported completing a bachelor’s degree, which is somewhat higher than the national average. Compared with national statistics, the distribution of household income was somewhat skewed toward the higher end. The median number of child-related services per week reported by parents was 8 h of public services and 3 h of private services.
Demographics for parents with and without a child with ASD.
ASD: autism spectrum disorder; HS: high school.
Of the baseline sample of 312 parents, 176 (56%) completed more than 50% of the 1-week follow-up survey. The demographic characteristics of these samples did not differ from the baseline sample, with the exception of income, as depicted in the Supplementary material.
Parents of children without ASD
Parents who did not have a child with ASD were recruited through a targeted, boosted Facebook advertisement that ran for a total of 3 weeks and reached almost 102,000 individuals and had 3000 engagements. The advertisement targeted parents of children aged 6 years or younger and with an income less than US$75,000 for the last week to increase the income diversity of the sample. From this recruitment effort, 1757 individuals initiated the survey. The 698 parents who met eligibility criteria, consented to being in the study, and completed 90% of the baseline survey lived in 696 zip codes. In relation to the most recent US census data, the percentage of parents reporting their race as White was about 10% higher and those reporting their race as Black or African American was about 5% less. The percentage of Hispanic/Latino participants was approximately equal to national estimates. Participants represented a broad range of educational background, about 50% having completed a bachelor’s degree, somewhat higher than the national sample. Compared with national statistics, the distribution of household income was somewhat skewed toward the higher end. The median number of services per week was 8 h of public services and 3 h of private services. As demonstrated in Table 1, these parents were less diverse than those of children with ASD: chi-square analyses revealed that significantly more parents of children without ASD were female (93.4%) and White (85.7%) and fewer participants were of Hispanic/Latino ethnicity (5.7%). In addition, the analyses showed that parents of children without ASD tended to be more highly educated (75% reported completing at least a bachelor’s degree) and reported higher income (67% reported annual income of US$75,000 or greater). The children of this sample were evenly distributed among males and females, and their median age was 3 years. Fifteen of the participants indicated that at least one of their children had another medical condition, developmental disability, or mental health problem other than autism. Parents of children without ASD reported an average of less than 1 h of either public or private service hours per week. Of the 698 individuals who completed the baseline survey, 334 completed more than 50% of the 1-week follow-up and were included in the analyses before and after viewing the See Amazing materials. The demographic characteristics of the 334 parents who completed more than 50% of the 1-week survey and were included in the follow-up analyses were similar to the baseline sample, with the exception of a marginally significant difference in gender, as depicted in the Supplementary material.
Participants’ data were excluded if they completed less than 50% of the study at either the Time 1 pre-test (n = 0) or the Time 2 post-test (n = 12). For the pre-test, 230 participants (20 males, 209 females, 1 “gender queer”) completed the implicit measure and the questionnaires. For the post-test, 115 participants (11 male and 104 female) completed the implicit measure and the questionnaires. Sixty-nine participants (6 male and 63 female) completed all measures at both time points. Due to the small number of males at each time point, all analyses below were conducted with female participants only. Approximately, half of this group had either heard of the website or had some familiarity with it. Approximately 40% of parents had no knowledge of the site and only 4% had viewed the entire site.
Procedure
Potential participants completed an initial set of screening questions regarding their child’s age and whether their child was autistic. If their responses were appropriate for either the group of parents with or without a child with ASD, participants were directed to an informed consent form and then administered the baseline measures. Participants were then sent an email asking them to spend 1.5–2 h reviewing the See Amazing videos and other web content. One week later, an email with the link to the post-baseline measures was sent to participants. The post-test included the same measures collected at baseline. This study was approved by the Institutional Review Boards at Georgetown University, Children’s National Health System, and William and Mary.
Materials
Website
The See Amazing website included general information about autism and links to resources for families and the public (http://autism.sesamestreet.org/). The website featured the initiative’s signature ‘The Amazing Song’ Music Video, 19 narrated videos that reflect a diverse group of children with ASD, an electronic storybook designed to be read by parents and children together that features Julia, a Muppet character with autism, and eight daily routine card sets. The routine cards are designed to be used by parents with their child to build new independent living and community skills, such as brushing their teeth or crossing the street.
Measures
Implicit measure
The IAT (Greenwald et al., 1998) assesses unconscious attitudes that individuals have about two groups or targets and two attributes (i.e. pleasant, unpleasant). The difference in reaction times to trials pairing one group and one attribute compared with reaction times (RTs) in trials pairing the other group with the other attribute is thought to reflect the strength of the association between the first and the second pair (Greenwald et al., 1998; Teachman & Woody, 2003). The IAT was administered as part of each session’s measures. To complete the task, participants were re-directed via a URL to complete the IAT using Inquisit software (www.millisecond.com), which necessitated downloading the program onto participants’ computers. A pilot test was conducted to select appropriate stimulus words for adapting the IAT to examine attitudes toward autistic individuals. In the pilot study, 100 adults (Mage = 32.3; SDage = 9.1; 35% female) from the United States participated through Amazon’s Mechanical Turk, an online crowdsourcing data collection site that provides easy access to a large, stable, and diverse subject pool (Mason & Suri, 2012). Adults were asked to generate as many words as they could to describe “young children with autism” and “young children without autism.” The order in which participants answered these two free response questions was counterbalanced. Research assistants identified the words that were most frequently used for one category, but infrequently used for the other category. Words that received the highest values on one category and the lowest values for the other were chosen. The co-authors then selected descriptive words that were associated with one group, but not the other (autistic, spectrum, ASD, Asperger’s, typical, non-autistic, usual, non-autistic). These words were then incorporated into the IAT for the categories non-autistic children and autistic children. Participants were given the category that each word was associated with prior to the task. The words associated with pleasant and unpleasant categories were the standard IAT words (e.g., marvelous, superb, pleasure, joyful, beautiful, glorious, horrible, awful, tragic, agony, painful, terrible).
The first two blocks of the IAT consisted of 24 trials each, with participants categorizing autistic words and non-autistic words with one response key on a keyboard, and “good” and “bad” words with another key. All stimulus words were presented in the middle of the screen with the category words presented in the top right and top left portions of the screen. In the second two blocks, which presented 64 trials, participants grouped all words into one of two categories, using a response key designated for each: half of the participants were randomly assigned to have good/autistic children and bad/non-autistic children as the two response options and half were randomly assigned to have bad/autistic children and good/non-autistic children. In the next block of 24 trials, the response keys for bad and good were switched. Finally, another 64 trials were presented in two subsequent blocks, with participants grouping all words into the other pairing of categories that were not previously presented. For each trial, the words were presented until participants responded. If they responded incorrectly, a red “X” appeared on the screen until they made the correct response. After participants made a correct response, there was a 250 ms inter-trial interval. The order in which these blocks were presented was counterbalanced across participants.
Survey
Services
To assess the public and private services that participants’ children received, parents were asked to indicate, on average, how many hours per week their child received public services (examples provided: special education classroom, speech/language therapy, occupational therapy, etc.) and private services (examples provided: paid for through insurance or out of pocket; including speech/language, occupational therapy, applied behavior analysis (ABA), counseling, etc.).
Acceptance of ASD
Participants viewed a video depicting a child with autism. To measure participants’ attitudes toward the child with autism portrayed in the video, they were provided with seven statements about the child and asked to rate each based on the following scale: 1 = strongly disagree, 2 = disagree, 3 = don’t know, 4 = agree, 5 = strongly agree (Harnum et al., 2007). Higher scores indicated more positive attitudes toward the child with autism. Example statements were “The child in the video is as smart as other young kids” and “I would feel comfortable spending time with this child.”
Knowledge of ASD
Participants’ knowledge about children with autism was assessed with 22 items. For six of these items, participants checked the statements that they thought were things that can be more difficult for children with autism compared with other children, including looking people in the eye, joining into play with others, and playing with a big group of children. The remaining items asked participants to indicate which statements are true about “children with autism” (13 items), including “they are not as smart as other kids” and “they can’t feel love,” and statements about autism in general (3 items), including “it is caused by bad parenting” and “it can help parents of children with autism to talk with one another” (reverse coded). The maximum score was 22 points, with higher scores indicating better knowledge about autism.
Familiarity
To assess familiarity with the website, participants were asked, “Are you familiar with the Sesame Street See Amazing in all Children resources?” and were given the following choices: “No,” “Yes, I’ve heard of it,” “Yes, I’ve seen some of the website,” “Yes, I’ve seen the entire website.”
Parenting competence
Family empowerment was measured with the 8-item competence subscale of the Family Empowerment Scale (FES; Singh et al., 1995) that asks how frequently they have certain feelings and beliefs related to different areas of parenting. The instructions indicated that the questions included different activities that parents/guardians may or may not do. In addition, they were instructed to answer the questions by thinking of their own situation only, and not the situations of others. The response scale was a 5-point Likert-type scale (1 = never, 2 = seldom, 3 = sometimes, 4 = often, 5 = very often). An example item is “I feel confident in my ability to help my child grow and develop.” Higher scores indicated greater feelings of empowerment.
Parenting confidence
Parents’ confidence in helping their child with daily routines and other activities in the community was assessed with two items (e.g., “I feel confident in my ability to help my child through daily routines; e.g., morning routine, bedtime routine, brushing their teeth, washing their hands” and “I feel confident in my ability to help my child through other activities; e.g., going to a restaurant, going to the store, crossing the street, trying a new food”). These two items were summed and averaged to create the parenting confidence score.
Parenting strain
The degree to which parents view their child’s behavior as stressful was measured through the Revised Short Form of the Caregiver Strain Questionnaire (CGSQRvSF; Brannan et al., 1997). Participants were asked to think back over the past month and indicate how much of the following problems has been a result of their child’s problems: interruption of personal time, missing work or neglecting other duties, financial strain, disruption of family relationships, feeling sad or unhappy, feeling worried about child’s future, and feeling tired or strained. Response choices to the seven questions were 1 = not at all, 2 = a little, 3 = somewhat, 4 = quite a bit, 5 = very much. Higher scores indicated more strain.
Parental stigma
Parental stigma was measured with items adapted from the Affiliate Stigma Scale (Mak & Cheung, 2008; Werner & Shulman, 2015). The Social Engagement subscale consisted of five items to examine parents’ comfort with involving their child in community activities (e.g., “I am comfortable explaining to other people how to best include my child”). Parents were asked to indicate their level of agreement on a 5-point scale, stretching from 1 = strongly disagree to 5 = strongly agree. The Affective Stigma subscale was only presented to parents of children with autism, tapping feelings of rejection and bias that a parent might experience as a result of their child’s diagnosis (e.g., “My child’s behavior makes me feel embarrassed”). Responses ranged from 1 = strongly disagree to 5 = strongly agree. Higher scores indicated greater feelings of stigma.
Demographics
Participants also provided their gender, ethnicity, race, level of schooling completed, and income.
Data analysis
The IAT was scored using Greenwald et al.’s (2003) revised method. Data from the blocks of interest (i.e., those displaying both labels (e.g., autistic and good; non-autistic and bad) were used in the calculations. The first two blocks of interest contained labels with the first pairing, which were autistic-bad and non-autistic-good or autistic-good and non-autistic-bad, depending on the counterbalanced condition to which the participant was randomly assigned. Blocks of interest 3 and 4 contained labels with the opposite pairing. Trials with RTs greater than 10,000 ms were deleted to reject trials on which participants are not attending to the task. Incorrect trials were replaced with the mean of the block plus 600 ms (see Greenwald et al., 2003). A standard deviation for the latencies for the trials in the first and the third blocks of interest were pooled; the same was done for the latencies for the trials in the second and the fourth blocks of interest. Difference scores were then computed subtracting the average latency of the third block of interest from that of the first block of interest. An additional difference score was computed subtracting the average latency of the fourth block from that of the second block of interest. Each of these two difference scores was then divided by the pooled standard deviations for the respective blocks. These two resulting scores were averaged; this number was the final d score. Positive d scores indicated a stronger implicit bias against children with autism.
Results
Parental implicit bias
Pre-test
For the pre-test, the mean calculated d score was 0.37 (SD = 0.47, range −0.98 to 1.42), which significantly differed from 0, t(208) = 11.41, p < 0.001, indicating an overall implicit bias against children with autism. To examine differences in implicit bias between parents of children with autism and without autism, an independent t-test was conducted. At pre-test, results revealed that parents of children without autism (n = 116; M = 0.54, SD = 0.37) had more biased attitudes than parents of children with autism (n = 93; M = 0.15, SD = 0.48), t(207) = –6.62, p < 0.001. The mean of the parents of children without autism was significantly different from 0, t(115) = 15.75, p < 0.001, as was the mean of parents of children with autism, t(92) = 3.01, p = 0.003, suggesting that both groups showed an overall implicit bias against children with autism.
Post-test
For the post-test, the mean calculated d score was 0.23 (SD = 0.46, range −0.85 to 1.08), which significantly differed from 0, t(102) = 5.21, p < 0.001, indicating an overall implicit bias against children with autism. At post-test, parents of children without autism (n = 38; M = 0.33, SD = 0.47) showed no difference in implicit attitudes compared with parents of children with autism (n = 65; M = 0.18, SD = 0.44), t(101) = –1.68, p = 0.096. The mean d score of the parents of children without autism was significantly different from 0, t(37) = 4.30, p < 0.001, as was the mean of parents of children with autism, t(64) = 3.01, p = 0.004, suggesting that both groups showed an implicit bias against children with autism in the post-test.
Pre-test versus post-test comparison
To examine whether implicit bias changed from the pre- to the post-test, a paired samples t-test was conducted for participants who completed the IAT at both time points. Results indicated that, overall, there was no difference in implicit bias between the pre- and post-tests, t(61) = −0.16, p = 0.877. When these analyses were conducted separately by group, there were also no differences for parents of children without ASD, t(39) = –1.02, p = 0.314, nor for parents of children with ASD, t(21) = 1.06, p = 0.303.
To examine whether a change in implicit bias was related to any of the variables of interest in the pre-test, a regression analysis was conducted with IAT change as the dependent variable. Parental group as well as pre-test acceptance, knowledge, familiarity, parenting competence, strain, and stigma (both subscales) were used as predictor variables. The overall model was significant, F(7, 54) = 2.30, p = 0.042, R2 = 0.26. Results demonstrated that people of a child without autism showed a greater reduction in implicit bias than parents of a child with autism, β = −0.46, p = 0.034. In addition, the more negative participants’ initial explicit attitude toward autism, the greater reduction in implicit bias, β = 0.53, p = 0.007. Results also revealed that the less knowledge participants had at the pre-test, the greater reduction in implicit bias, β = −0.39, p = 0.059, although this analysis did not reach statistical significance.
Predictors of competence in parents of children with ASD
To examine whether changes in parental competence were predicted by changes in implicit bias, acceptance, knowledge, strain, and stigma (Affective Stigma subscale), regression analyses were conducted with pre-post differences in parental competence as the dependent variable with pre-post difference scores of the other variables as predictors. Results indicated that changes in parental competence was associated with changes in attitudes, β = 0.57, p = 0.012, strain, β = −0.43, p = 0.033, and knowledge, β = −0.76, p = 0.001. No other variables yielded significant effects.
Discussion
The present study sought to characterize implicit bias toward children with autism in parents of children with autism and those of children without autism, assess the effects of the See Amazing Sesame Street online autism resources on these two different samples, and examine whether attitude change and knowledge about autism would be associated with greater self-reported competence in parents of children with autism. Hypotheses were partially supported. These results have implications for improving attitudes toward autistic children as well as providing positive outcomes for parents of children with ASD.
At baseline, both parents of children with ASD and those of children without ASD showed significant negative implicit bias against children with autism. These findings are consistent with other reports showing that teachers and tutors display negative implicit attitudes toward children with autism (Kelly & Barnes-Holmes, 2013). In the current study, parents of children without ASD showed significantly greater implicit bias toward children with autism compared with parents of children with autism at baseline. This finding may be attributed to different levels of knowledge between the parent groups; indeed, more knowledge about the See Amazing resource was associated with less implicit bias.
At the post-test, there was no longer a statistically significant difference in implicit bias toward children with autism between parents with and without a child with autism; parents of children without autism showed a greater reduction in implicit bias compared with those of a child with autism. One explanation for these findings may be that parents of children with autism already have considerable knowledge and experience with the characteristics of autism and thus may not learn as much from the online See Amazing resources. On the other hand, those parents without a child with ASD may have gained greater knowledge from the online resources than parents of a child with ASD. This conclusion is partially supported by the result that those participants with less knowledge showed a greater reduction in implicit bias, although the result did not reach statistical significance (p = 0.059). In addition, those with more negative explicit attitudes toward children with autism in the pre-test tended to show a greater reduction in implicit bias. This finding is in line with previous research by Mahoney (2008) that suggests that individuals with more knowledge of autism report more explicit positive attitudes toward individuals with autism. Together, our findings suggest that the online educational resources used in the current study provided a view of autism that could lead to a reduction in implicit bias against children with autism.
An additional goal of this study was to examine whether improving attitudes toward and knowledge about children with ASD may mitigate some of the psychological issues associated with parenting children with ASD. Specifically, we tested whether a positive change in attitudes and knowledge would increase feelings of empowerment (i.e., feelings of parental competence) in parents of children with ASD. Results partially supported this hypothesis, indicating that a more positive change in explicit attitudes and increased knowledge from the pre- to the post-test was associated with a greater change in empowerment in parents of children with ASD. These results are consistent with the FAAR model (Patterson, 1988) which argues the importance of resources for parents of ASD children. As greater feelings of empowerment are associated with higher use of services, more awareness of support resources, and higher self-efficacy (Wakimizu et al., 2011), the current research suggests that exposure to the See Amazing website may have some positive consequences for parents of children with ASD. Although not the focus of the current study, increased empowerment may also be associated with less negative psychological outcomes; indeed, we demonstrated that positive changes in empowerment were associated with less strain from the pre- to the post-test in parents of children with ASD.
Although changes in explicit attitudes and increased knowledge were associated with greater empowerment in parents of children with ASD, our hypothesis regarding implicit attitude change was not supported. That is, results from the current study demonstrated that changes in implicit bias against children with autism were unrelated to feelings of empowerment at the post-test. One explanation for the dissociation of results is that implicit and explicit attitudes are separate constructs; research generally reveals a small correlation or no correlation between explicit and implicit measures (Flood et al., 2013; Hofmann et al., 2005; Karpinski & Hilton, 2001; Oswald et al., 2013; Thomas et al., 2007). Each type of measure predicts different behaviors, which underscores the importance of assessing both of these constructs.
Some caveats need to be acknowledged for interpreting the present results. In particular, only 50% of the participants who completed the IAT at the pre-test also completed the IAT at the post-test. We cannot rule out the possibility that there are characteristics of those participants who only completed the IAT during the pre-test session that would have led to a different pattern of results if these participants had completed the post-test IAT session, although the demographic characteristics of each group did not differ from the pre- to the post-test. In addition, such a large decrease in sample size may have limited our statistical power to detect reliable associations between the variables in the present study. Another potential limitation of the current study is that the IAT was administered online as opposed to in a controlled laboratory setting. Thus, participants could have been completing the task in a distracting situation. Previous research, however, has shown that IAT findings do not vary as a function of setting (i.e., online vs lab; Houben & Wiers, 2008). Finally, the parents of children with autism were primarily recruited through online advocacy groups, and thus the findings regarding this group of parents may not generalize to other groups of parents.
The present results provide preliminary evidence that there is a relationship between exposure to the Sesame Street online resources and changes in implicit attitudes toward children with autism for some individuals. These findings suggest that those with less familiarity with autism, namely parents of children without autism; those with more negative initial attitudes toward children with autism; and those with less knowledge about autism differentially benefit from the online resources. Although more research on the online resources needs to be conducted, our results suggest that a brief (i.e., 2 h) interaction with the website, which is associated with a familiar brand (i.e., Sesame Street), may reduce bias toward children with autism. These findings are especially important in light of the fact that advocacy efforts to increase positive attitudes toward autism have recommended more interaction between non-disabled individuals and those with disabilities (Metzel & Walker, 2001), based on previous work demonstrating that contact with individuals with disabilities can reduce prejudice in non-disabled individuals (Keith et al., 2015; Nevill & White, 2011). Results from the current study, however, suggest that these interactions could potentially be harmful to children with autism if they are exposed to social situations in which they interact with people holding biases toward their group. Thus, acquiring knowledge from a website may serve as an easy, quick way to reduce bias without potential harmful consequences toward individuals with autism. To extend this work, future research directions include developing a better understanding about which aspects of the Sesame Street online resources are most beneficial in changing implicit attitudes. Researchers can also examine how to best recruit adults with social distance from autism and those who do not have children to look at these online resources to reduce potential biases against children with autism. In addition, future research is needed to better understand how changes in implicit attitudes lead to differences in behavior toward children with autism. Our results provide an encouraging step in improving attitudes toward individuals with autism.
Supplemental Material
Supplementary_Materials_Revision_Autism – Supplemental material for Assessment of Sesame Street online autism resources: Impacts on parental implicit and explicit attitudes toward children with autism
Supplemental material, Supplementary_Materials_Revision_Autism for Assessment of Sesame Street online autism resources: Impacts on parental implicit and explicit attitudes toward children with autism by Cheryl L Dickter, Joshua A Burk, Laura Gutermuth Anthony, Hillary A Robertson, Alyssa Verbalis, Sydney Seese, Yetta Myrick and Bruno J Anthony in Autism
Supplemental Material
Supplementary_Materials_Revision_Autism_Correlation_Table – Supplemental material for Assessment of Sesame Street online autism resources: Impacts on parental implicit and explicit attitudes toward children with autism
Supplemental material, Supplementary_Materials_Revision_Autism_Correlation_Table for Assessment of Sesame Street online autism resources: Impacts on parental implicit and explicit attitudes toward children with autism by Cheryl L Dickter, Joshua A Burk, Laura Gutermuth Anthony, Hillary A Robertson, Alyssa Verbalis, Sydney Seese, Yetta Myrick and Bruno J Anthony in Autism
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Sesame Workshop (Grant/Award No. “GR411457”).
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
