Piecing together the puzzle: development of theSocietal Attitudes towards Autism (SATA) scalejrs3_1224 1..8Luci N. Flood, Amanda Bulgrin and Betsy L. MorganUniversity of Wisconsin La Crosse, USA
Key words: Autism, attitudes, scale development, college students.
The rise in the prevalence of autism creates a needfor a reliable and valid measure of attitudes towardsautism. The current study describes the develop-ment of a brief 16- item measure of Societal Atti-tudes towards Autism (SATA) that exhibits soundpsychometric properties and has a demonstratedability to discriminate between expert and generalcollege student samples. The final SATA was theresult of pilot work on 75 items and exploratory andconfirmatory factor analyses on a 45-item versionwith 475 undergraduates. Knowledge and personaldistance subscales yielded inconsistent reliabilityand validity outcomes. The SATA showed strongcontent and construct validity as evidenced byknown groups discrimination, and predicted asso-ciations with an attitude towards disability measure,an autism preference item, and a measure ofimplicit attitudes towards disabilities.
Autism is a disorder that affects more children than juvenilediabetes, childhood cancer, and paediatric AIDS combinedeach year. Experts estimate that autism is diagnosed in oneout of every 91110 births and affects 1.5 million people inthe USA (Kogan et al., 2009; National Survey of ChildrensHealth, 2007). Autism is a general term used to describe agroup of complex developmental brain disorders known asPervasive Developmental Disorders (American PsychiatricAssociation, 2000). More specifically, autism spectrumdisorders are characterised by atypical development insocialization, communication, and behavior, and symp-toms generally include abnormalities in cognitive func-tioning, learning, attention, and sensory processing (Rice,2009, p. 1). Autism encompasses a complex spectrum ofbehaviours and outcomes that are reflected in a commonsaying in the autism community, If youve met one personwith Autism youve met one person with Autism (Shore,Rastelli and Grandin, 2006).
The rise of autism rates has been linked primarily toa combination of increased awareness and more accuratediagnoses (e.g., Hertz-Picciotto and Delwiche, 2009).Given the large number of potential children and adults withautism diagnoses now and in the future, it is important toensure that the general public is educated about autism andprepared to interact with people with autism. In addition,
due to the American Disabilities Act, current students aremore likely to interact with high-functioning individualswith autism across several educational settings. Nevilland White (2011) found that college students with a firstdegree relative with autism reported more openness towardsinteracting with individuals with autism than did collegestudents without such experience.
Despite the high levels of autism diagnosis, attitudestowards autism are relatively understudied. One aspect ofunderstanding current views towards autism is to developmeasures to accurately detect individuals authentic atti-tudes towards autism. An accurate measure will giveresearchers and practitioners a better understanding of thepublics orientation towards autism. This paper describesthe revision and validation of a measure designed specifi-cally to assess attitudes towards autism.
Attitudes towards disabilitiesSocial scientists describe attitudes as the psychologicaltendency to evaluate people or beliefs as either favourableor unfavourable (Eagly and Chaiken, 1993). Much of pastresearch on attitudes towards disabilities has focused onphysical disabilities (e.g., Rojahn, Komelasky and Man,2008; Stewart, 1988). In addition, past measures tend tomeasure attitudes towards disabilities at a more global level(across multiple disabilities) rather than specific level(a particular disability or diagnosis) (Antonak, 1981; Pruettand Chan, 2006; Seo and Chen, 2009; Yuker, Block andYoung, 1966).
Several measures exist to assess attitudes towards autismfor specific subgroups of individuals. Many of the existingmeasures involve reactions to the behaviour of childrenwith autism by adults (Iobst, Nabors and Rosenzweig et al.,2009), parents (e.g., Hebert and Koulouglioti, 2010), teach-ers (e.g., Park, Chitiyo and Choi, 2010) and schoolmates(e.g., Campbell, 2008; Morton and Campbell, 2008; Reiterand Vitani, 2007; Silton, 2010). In addition, several studieshave focused on the relationship between attitude andknowledge of autism on the placement of children withautism in mainstream classrooms (e.g., Horrocks, Whiteand Roberts, 2008; Middleton, 2006; Simpson, Griswoldand Myles, 1999). Fewer studies exist with the purpose ofproviding an attitudes towards autism measure aimed at amore general population.
Journal of Research in Special Educational Needs Volume Number 2012 doi: 10.1111/j.1471-3802.2011.01224.x
1 2012 The Authors. Journal of Research in Special Educational Needs 2012 NASEN. Published by Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and350 Main Street, Malden, MA 02148, USA
Mahoney (2008) adapted the Mental Retardation AttitudeInventory- Revised (Antonak and Harth, 1994) to developthe Autism Attitudes Inventory (AAI). The AAI appears tobe the first scale specifically developed to measure attitudestowards autism and includes approximately 60 questionsregarding interactions with people with autism that are cat-egorised into three major areas: social distance (the extentto which respondents are willing to interface with personswith autism across situations), academic integration (beliefsregarding inclusion of children with autism in mainstreamschools) and private rights (the rights of individuals withautism). Although one of the first serious attempts at anattitudes towards autism scale, Mahoneys scale does notinclude items that tap respondents knowledge aboutautisms causes and behavioural outcomes, has not beenfully vetted for construct validity and has not been pub-lished in a peer-reviewed journal. Given the large variabilityin the publics beliefs about the causes of autism (e.g.,Goin-Kochel and Myers, 2005), some of which are errone-ous (e.g., vaccines, gluten-free diets) and the relativelysmall number of consistently apparent behaviours associ-ated with autism behaviours, we felt that the addition ofknowledge items was important in an attitudes towardsautism scale.
Developing an attitudes towards autism scaleIn this paper, we describe the development of an attitudetowards autism scale and report on the processes associatedwith establishing its reliability and validity. The purpose ofthe study was to develop an attitude towards autism scalethat reflected societal attitudes and knowledge of the disor-der. The end result is a 16-item measure that can be used toassess societal attitudes towards autism. The developmentof a psychometrically sound measure assessing attitudestowards autism services several functions. An attitudestowards autism scale allows researchers to distinguishbetween more general attitudes towards disabilities and spe-cific attitudes towards autism. A psychometrically soundscale allows researchers to assess change in attitudestowards autism over time and/or between groups. In addi-tion, the relationships between predictors and correlates ofattitudes towards autism can be explored. Finally, an atti-tudes towards autism scale can be used as an educationaland/or screening tool to enhance dialogues with pre-serviceteachers and other individuals who will be working directlywith children or adults with autism. Overall, scholarly pur-suits, social policy and educational ventures regardingautism may all be informed by the appropriate use of anattitudes towards autism scale.
MethodThe development of an Attitudes towards Autism Scale(ATA) was comprised of several stages. The original devel-opment of items built upon the 23 attitudes towards autismitems from Mahoneys (2008) dissertation research result-ing in a 23-item scale representing three factors socialdistance, academic integration and private rights. We modi-fied and added items according to his three subscales forvalidation with a younger participant pool. For instance,Mahoney had an item reading Willing for my child to have
a child with autism as a close friend that we modified tobe more salient to college students reading I would becomfortable having a friend with autism. In addition, wecreated 35 items developed for a potential subscale createdto assess knowledge regarding autism behaviours andcauses. Many of the added items were based on thetwo primary authors one-on-one work with children withautism via community-service organisations. As indicatedin the results, four of Mahoneys original items (three ofwhich were modified) remain in the final scale.
A pilot sample of 54 undergraduates completed the firstversion of the scale created from Mahoneys work with ourmodifications and additions and comprised of 75 items.This pilot group received extra credit for participation, com-pleted the scale online, and was provided textboxes afterevery set of five items with the prompt Please review thequestions above. Are any of them confusing? Poorlyworded? Not make sense to you? If so, please indicatewhich question by number and explain the problem. Inaddition, a response item of not enough knowledge wasincluded with each item.
The second version of the ATA was comprised of 45 itemsculled from the original 75 after discarding items thatshowed low variance, were flagged as vague and/or a sub-stantial percentage of respondents answered not enoughknowledge. In addition, items were modified based on thefeedback from two autism experts (a college professor anda supervisor at a local institution for children with cognitivedisabilities). For example, an item in the knowledge sub-scale previously worded Individuals with autism displayflapping behaviors was reworded to read All individualswith autism demonstrate repetitive behaviors, such asrocking or flapping of arms or hands. The 45-item versioncontinued to represent the four hypothesised subscales:social distance, academic integration, private rights andknowledge of behaviour and causes.
Four hundred and seventy-five US undergraduates (70%female; 90% Caucasian, representing a wide range ofmajors from across the campus) completed the 45-itemscale online and received extra credit for participation(2 credits provided for a class with approximately 700points available from traditional assessments such asexams, discussions and worksheets). Unfortunately, 182participants failed to provide demographic information,although they completed the scale items. All versions of theATA utilised a four-point response scale of 1 strongly dis-agree to 4 strongly agree with no midpoint and coded suchthat a high score indicates high acceptance or high knowl-edge. In addition, respondents provided the extent to whichthey have had contact with individuals with autism, whichwas coded as 1 experience for each of the following: Littleto no contact, have been employed somewhere where Iwork with a person with autism, have a friend with autism,have a family member with autism, have volunteered witha person with autism, have been a primary caregiver for aperson with autism, other (Range 04; M = 0.77; SD = 0.76;
Journal of Research in Special Educational Needs,
2 2012 The Authors. Journal of Research in Special Educational Needs 2012 NASEN
39% of the sample had no contact with an individual withautism).
Validity measuresIn addition to completing the ATA, the 475-undergraduatesample also completed several measures aimed at establish-ing the validity of the ATA; several of the validity measureswere collected with the original sample but several weeksafter the initial ATA data collection.
Preference towards persons with autism. At the time of theATA data collection, respondents also completed a seven-point item measuring their overall appraisal of individualswith autism ranging from I strongly prefer people withautism to people without autism to I strongly prefer peoplewithout autism to people with autism M = 4.89; SD = 0.94indicating a slight preference for people without autism topeople with autism.
Attitudes towards disabled persons. Two weeks aftercompleting the second version of the ATA, a subset of theundergraduate sample (n = 283) completed the Attitudestowards Disabled Persons (ATDP) online for additionalextra credit points. The ATDP measures attitudes towardspeople with disabilities via respondents self-report and isone of the most highly utilised scales in the field (Yuker,Block and Young, 1970). The ATDP is valid and reliableand somewhat susceptible to social desirability (Pruettand Chan, 2006). The ATDP has three different formats:ATDP-O, ATDP-A, ATDP-B. Form O and A were found tobe the most reliable and valid. In order to shorten the scale,we selected 14 questions most germane to the current studyfrom forms O (questions 2, 4, 6, 7, 13 and 17) and A(questions 3, 5, 6, 7, 8, 24, 26 and 29). Sample questionsincluded People with disabilities are just as intelligent aspeople without disabilities and People with disabilities aremore emotional than people without disabilities. All of theitems were reworded to use people first language fromdisabled persons to people with disabilities.
Behavioural intention. At the time of the ATDP adminis-tration, students were asked a question aimed at assessingtheir behavioural intention towards interacting with indi-viduals with disabilities. Would you be interested in havingyour name and email provided to a local community orga-nization that utilizes students to work with individuals withdisabilities for potential contact in the Spring regardingvolunteering? Four options were coded: 1 = No I am notinterested (26%), 2 = No not at this time (48%), 3 =Yes but not for this coming semester (5%), or 4 = Yes please submit my name and email (20%). The responsesfrom the students were matched to earlier responses by aunique identifier. The survey was designed such that stu-dents who indicated that they wished to have their contactinformation provided to a volunteer coordinator providedthis information in an independent survey. Consequently,the names and emails from the students were separatedfrom all other responses. The contact information of stu-dents who indicated 3 or 4 and provided contact infor-mation was provided to a local direct service community
organisation as a list of students who were potentiallyinterested in volunteering. We do not know if the organisa-tion utilised the information and/or if any of the studentsprovided service to the organisation.
Disability attitudes implicit association test. At the con-clusion of the survey containing the ATA items, participantsindicated their interest in taking part in a related study.Approximately 8 weeks after the survey, 58 respondentsreceived additional extra credit for completing an implicitassociation test...