Autism refers to a neurodevelopmental condition, characterised by differences in communication and social interaction in conjunction with repetitive behaviours, activities, or interests and hyper- or hyposensory sensitivities (American Psychiatric Association, 2013). According to the United States Centers for Disease Control and Prevention, the prevalence of autism in children is 2.3% (Maenner et al., Reference Maenner, Shaw, Bakian, Bilder, Durkin, Esler, Furnier, Hallas, Hall-Lande, Hudson, Hughes, Patrick, Pierce, Poynter, Salinas, Shenouda, Vehorn, Warren, Constantino and Cogswell2021). Similar prevalence rates of over 3% were reported among Australian children (Australian Bureau of Statistics, 2019). Although this prevalence of autism has created greater interest and research in the field, studies focusing specifically on gender differences and the experiences of females on the autism spectrum are still new and developing (Gould, Reference Gould2017).
The male to female diagnosis ratio for autism has been reported to be between 4:1 to 3:1 (Baio et al., Reference Baio, Wiggins, Christensen, Maenner, Daniels, Warren, Kurzius-Spencer, Zahorodny, Rosenberg, White, Durkin, Imm, Nikolaou, Yeargin-Allsopp, Lee, Harrington, Lopez, Fitzgerald, Hewitt and Dowling2018; Loomes et al., Reference Loomes, Hull and Mandy2017). Over the past decade, however, there has been increased recognition that females on the autism spectrum are more likely to be under-identified, misdiagnosed, or diagnosed later in life than males on the autism spectrum (Begeer et al., Reference Begeer, Mandell, Wijnker-Holmes, Venderbosch, Rem, Stekelenburg and Koot2013; Loomes et al., Reference Loomes, Hull and Mandy2017). These diagnostic-related issues appear to be because of differences in the presentation of autistic characteristics from males on the autism spectrum (Bargiela et al., Reference Bargiela, Steward and Mandy2016; Green et al., Reference Green, Travers, Howe and McDougle2019). A variety of theories, including that of extreme male brain theory (Baron-Cohen, Reference Baron-Cohen2002) and female protective effect (Robinson et al., Reference Robinson, Lichtenstein, Anckarsäter, Happé and Ronald2013), have sought to explain the differences between males and females on the autism spectrum, but there is no consensus currently.
Hiller et al.’s (Reference Hiller, Young and Weber2014) study with clinician and teacher diagnostic assessments and reports indicated that females on the autism spectrum were more able to engage in reciprocal conversation and had different restricted interests than those common with males. A summary of behavioural sex differences in Lai et al. (Reference Lai, Lombardo, Auyeung, Chakrabarti and Baron-Cohen2015) also showed females had more awareness of social interaction needs and tendencies towards perfectionism. Another notable difference is that females on the autism spectrum appear to mask or camouflage their autism characteristics, such as social communication difficulties, more than males on the autism spectrum (Lai et al., Reference Lai, Lombardo, Ruigrok, Chakrabarti, Auyeung, Szatmari, Happé and Baron-Cohen2017; Schuck et al., Reference Schuck, Flores and Fung2019). Although the terms masking and camouflaging are often used interchangeably, for simplicity we will only use the term masking in this review to refer to the complex behaviours of imitating others and concealing difficulties to meet environmental demands (Tubío-Fungueiriño et al., Reference Tubío-Fungueiriño, Cruz, Sampaio, Carracedo and Fernández-Prieto2021).
Tubío-Fungueiriño et al.’s (Reference Tubío-Fungueiriño, Cruz, Sampaio, Carracedo and Fernández-Prieto2021) systematic review explored the effect of masking in females on the autism spectrum. It identified that they masked through mimicking other’s behaviours, consciously developing their empathy skills, and exerting self-control. Tubío-Fungueiriño et al. proposed that masking could be connected to the social expectations of being a female and to better fit into social environments with reduced social rejection. It was also recognised that although masking served as an adaptive mechanism for females on the autism spectrum, there were substantial negative impacts, such as high anxiety and reduced self-esteem and wellbeing.
Peer Interactions of School-Aged Students on the Autism Spectrum
Peer interactions are critical to the development of children and adolescents, affecting their behavioural and affective wellbeing (Bukowski et al., Reference Bukowski, Laursen and Rubin2018). Peers are also a source of social support for neurotypical children and adolescents, and successful peer interactions portend adult adaptive and social functioning (Picci & Scherf, Reference Picci and Scherf2015). For individuals on the autism spectrum, their social difficulties extend into their interactions with peers at school. School-aged students on the autism spectrum reported difficulties with friendships, including concerns around not knowing how to initiate friendships and how to converse with friends (Sumiya et al., Reference Sumiya, Igarashi and Miyahara2018). They also encountered significant levels of victimisation and worries associated with victimisation (Ashburner et al., Reference Ashburner, Saggers, Campbell, Dillon-Wallace, Hwang, Carrington and Bobir2019; Sterzing et al., Reference Sterzing, Shattuck, Narendorf, Wagner and Cooper2012). The importance of peer interactions for school-aged students on the autism spectrum is reflected in the attention given to skills for enhancing peer interactions within evidence-based social skills and peer-mediated interventions (Babb et al., Reference Babb, Raulston, McNaughton, Lee and Weintraub2021; Bene et al., Reference Bene, Banda and Brown2014).
Children on the autism spectrum appeared to have fewer friends, less frequent contact with their friends, and shorter durations of friendship compared to their neurotypical peers (Petrina et al., Reference Petrina, Carter and Stephenson2014). Another finding in Petrina et al. (Reference Petrina, Carter and Stephenson2014) was that children on the autism spectrum were less able to provide a comprehensive definition of friendship, although they did identify companionship, similar interests and personality, and mutual help as part of friendships. Cresswell et al.’s (Reference Cresswell, Hinch and Cage2019) more recent review of peer interactions for adolescents on the autism spectrum underscored similar difficulties in describing friendship as a concept. Adolescents did, however, generally understand friendships and its important qualities. They also expressed a desire for friendship and were able to form and maintain friendships at school but encountered challenges in understanding social conventions, loneliness, and peer rejection.
Tomlinson et al.’s (Reference Tomlinson, Bond and Hebron2020) systematic review about the school experiences of girls on the autism spectrum found that they had a quality of friendship more similar to neurotypical girls than boys on the spectrum. They also often formed friendships with other girls on the autism spectrum and were more socially motivated than their male peers. It was also reported that girls on the autism spectrum found perspective taking and social expectations in friendships difficult. Moreover, girls on the autism spectrum faced relationally aggressive behaviours. These behaviours, which are defined as those that harm another’s social relationships or social group inclusion, include exclusion and gossip, which was not experienced by boys on the autism spectrum (Kraft & Mayeux, Reference Kraft and Mayeux2018).
It increasingly appears that the experiences of girls on the autism spectrum are not comparable to that of boys (Hiller et al., Reference Hiller, Young and Weber2014; Tubío-Fungueiriño et al., Reference Tubío-Fungueiriño, Cruz, Sampaio, Carracedo and Fernández-Prieto2021). Considering the under-representation and under-studied nature of females on the autism spectrum (Lai et al., Reference Lai, Lombardo, Auyeung, Chakrabarti and Baron-Cohen2015; Watkins et al., Reference Watkins, Zimmermann and Poling2014), it is particularly pivotal to hear from girls on the autism spectrum themselves rather than from other stakeholders, such as teachers or parents, in understanding their unique peer interaction experiences. Previous reviews have investigated peer interactions of autistic adolescents using qualitative evidence (Cresswell et al., Reference Cresswell, Hinch and Cage2019) and the broader school experiences of girls on the autism spectrum (Tomlinson et al., Reference Tomlinson, Bond and Hebron2020). In this scoping review, we intended to focus solely on the peer interactions of school-aged girls on the autism spectrum from their perspective through synthesising qualitative, quantitative, and mixed methods research. Consequently, the aim of this review was to explore the peer interactions of school-aged girls on the autism spectrum from their perspective. The research questions we sought to answer in this review are as follows:
-
1. What are the peer interaction experiences of school-aged girls on the autism spectrum?
-
2. How do the peer interaction experiences of school-aged girls on the autism spectrum compare to school-aged boys on the autism spectrum?
-
3. What peer interaction difficulties are experienced by school-aged girls on the spectrum?
-
4. What strategies did school-aged girls on the spectrum employ when experiencing peer interaction difficulties?
Methods
Search Strategy
A scoping review design was chosen to identify and collate research evidence (Arksey & O’Malley, Reference Arksey and O’Malley2005). The choice of a scoping review was based on its appropriateness in exploring emerging evidence of a broader scope as compared to that of a more specialised systematic review (Munn et al., Reference Munn, Peters, Stern, Tufanaru, McArthur and Aromataris2018).
The complete search process can be viewed in Figure 1. The combination of search terms used to search the four databases was (Autis*) OR (Asperger*) OR (ASD) OR (ASC) AND (Girl*) OR (Female*) OR (Gender) AND (Mainstream) OR (General education) OR (Special school) AND (Peer*) OR (Friend*) OR (Relationship*) OR (Social*) OR (Bully*) OR (Interact*) OR (Lone*). A manual journal search was also conducted in four international peer-reviewed journals whose focus is developmental disabilities. Additionally, an ancestral search was performed on the full texts identified to be included in the review. The above searches were completed between January and July 2020.
Figure 1. Flowchart of Search Process.
To be included, studies had to have participants who were school-aged girls on the autism spectrum in kindergarten to Year 12. The perspectives of other stakeholders, such as family members, could be present; however, their data would not be extracted because of the study’s aim to be solely from the perspective of girls on the autism spectrum. The participants had to have actively contributed to the data (e.g., interview, self-report surveys) and explicitly discussed their peer interactions. Studies that did not include the active participation of girls on the autism spectrum, such as observation-only studies, were excluded. Studies also had to be published between 2010 and 2020 and be peer-reviewed journal articles in English. The date range was chosen to begin from 2010, as research focusing on girls on the autism spectrum is a relatively new development.
Data Screening, Extraction, and Synthesis
A total of 314 records were identified through the database and manual journal search. After 124 duplicates were excluded, two authors independently screened the abstracts and then the full texts of each article. The interrater reliability measured by Cohen’s kappa was κ = 0.830 for the abstract screening and κ = 0.839 for the full-text screening. A further 173 records were excluded through the abstract and full-text screening, with 17 articles meeting inclusion criteria from the database and manual journal search. Four additional articles were identified through ancestral searches and were agreed on by two authors for inclusion, with the interrater reliability being κ = 0.642 for the abstract screening and κ = 0.75 for the full-text screening. Thus, a total of 21 articles were included for analysis in this review.
The main characteristics of the studies — that is, location, participants, aim, study design and key findings — were then extracted. These characteristics are presented in Table 1. Inductive content analysis was employed to systematically analyse the studies and describe the research phenomenon (Elo et al., Reference Elo, Kääriäinen, Kanste, Pölkki, Utriainen and Kyngäs2014). In the preparation stage, the first author prepared the data by selecting data relevant to the research aim. Thereafter, they coded the data and organised the codes into categories, subthemes, and themes. As other participants, such as boys on the autism spectrum, or stakeholders, such as family members, could be present in the studies, care was taken to draw data exclusively from girls on the autism spectrum. In cases where data was aggregated or synthesised in a way that it was not clear whether it could be correctly attributed to a girl or girls on the autism spectrum, it was not included. The second author checked the extracted data and coded data respectively for accuracy and applicability to the research aim. The three main themes and 15 subthemes developed through the data analysis are displayed in Table 2.
Table 1. Study Characteristics
Table 2. Theme and Subthemes
Quality Assessment
A quality assessment of the included studies was also conducted using the Standard Quality Assessment Criteria for Evaluating Primary Research Papers From a Variety of Fields (Kmet et al., Reference Kmet, Lee and Cook2004). The studies were assessed independently by the first and the second authors, and the authors resolved all disagreements by discussion. Quality assessment scores were interpreted by comparing percentages in which the relevant criteria were met (McGarty & Melville, Reference McGarty and Melville2018), with scores < 55% interpreted as weak, 55%–75% as moderate, and > 75% as strong. Out of the 21 articles included, 20 were strong and one was moderate.
All studies had identifiable designs that were appropriate in answering the stated research questions and conclusions supported by relevant results. Data analysis procedures were clearly described for most of the studies (14) and partially for some (seven). For studies where the qualitative assessment score was used (10), verification procedures to establish credibility were stated in seven studies and researcher reflexivity was discussed in eight studies. For studies where the quantitative assessment score was used (11), all studies defined measures with minimal potential for measurement bias and reported appropriate variance estimates.
Findings
Study Characteristics
The included studies were conducted in the United Kingdom (11), United States (four), Australia (three), Netherlands (one), Taiwan (one) and Belgium (one). The majority of studies collected participant data in schools (13), although some were collected at home (six) or other private locations (two). Studies differed in their design, with qualitative studies (nine) being more prevalent than quantitative studies (seven) or mixed methods studies (five).
Participant Characteristics
For simplicity, the term school-aged girls on the autism spectrum will hereafter be condensed to girls on the autism spectrum. Studies including girls and boys on the autism spectrum as participants (14) were more common than studies with girls on the autism spectrum only (seven). Participating girls all had a diagnosis of either autism spectrum disorder, autism, Asperger’s syndrome or pervasive development disorder not otherwise specified, apart from one student who was unable to be officially diagnosed due to cost. Co-occurring diagnoses included attention-deficit/hyperactivity disorder, oppositional defiant disorder, anxiety, depression, obsessive–compulsive disorder, learning disability, intellectual disability, speech and language difficulties, global developmental delay, dyspraxia, auditory processing difficulty, visual impairment, epilepsy, facial tic disorder, scoliosis, and gastro-oesophageal reflux disease.
Less than half the studies included stakeholders such as teachers and parents (eight). In total, there were 236 girls on the autism spectrum (5–20 years) out of 2,112 total participants. Student participants were enrolled in mainstream primary schools (three); mainstream secondary schools (six); special schools (two); a variety of educational settings (eight), including pupil referral units, alternative provisions, support units, supported employment and open university; and home education or not specified (two).
Themes and Subthemes
Friendship
Interest in friendship
Girls on the autism spectrum showed a keen interest for friendship (Cook et al., Reference Cook, Ogden and Winstone2018; Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Tierney et al., Reference Tierney, Burns and Kilbey2016). The participants in Tierney et al. (Reference Tierney, Burns and Kilbey2016) expressed their intrinsic motivation for friendship, with one student sharing, ‘I definitely couldn’t be some kind of hermit’ (p. 78). This is reinforced by the findings of Sedgewick et al. (Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016), where girls on the autism spectrum had higher levels of social motivation than boys on the autism spectrum. They also described a desire for a greater number of friends (Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016).
Additionally, girls on the autism spectrum reported loneliness from unfulfilled longings for friendship (Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). The absence of friends created feelings of anxiety and depression (Goodall & MacKenzie, Reference Goodall and MacKenzie2019) and made one feel ‘as though the whole world is against you’ (Vine Foggo & Webster, Reference Vine Foggo and Webster2017, p. 78). These accounts of loneliness are supported by the work of Bossaert et al. (Reference Bossaert, Colpin, Pijl and Petry2012), which highlighted that girls on the autism spectrum reported greater levels of loneliness than boys on the autism spectrum.
Girls on the autism spectrum valued the belonging that came from friendship (Myles et al., Reference Myles, Boyle and Richards2019; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). For the participants in Myles et al. (Reference Myles, Boyle and Richards2019), friendship was also seen to be integral to their overall school experience in providing belonging and social security. The notion of friends as social security was similarly referenced by Murphy et al. (Reference Murphy, Burns and Kilbey2017), wherein a student noted that having fewer friends could make one vulnerable to making enemies. Most participants in Vine Foggo and Webster (Reference Vine Foggo and Webster2017) stated it was critical to have friends and that the benefits of friendship included emotional support and contentment.
Despite an interest in friendship, some girls on the autism spectrum also expressed a preference to be alone sometimes (Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Hebron & Humphrey, Reference Hebron and Humphrey2014; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). They expressed pleasure in spending time on their own (Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Hebron & Humphrey, Reference Hebron and Humphrey2014) or needing solitude to de-stress (Vine Foggo & Webster, Reference Vine Foggo and Webster2017). The student in Goodall and MacKenzie (Reference Goodall and MacKenzie2019) who enjoyed her solitude also ranked having friends as being least important to an enjoyable school life.
Experiences of friendship
A considerable number of girls on the autism spectrum reported having friends or best friends (Calder et al., Reference Calder, Hill and Pellicano2013; Cook et al., Reference Cook, Ogden and Winstone2018; Dean et al., Reference Dean, Kasari, Shih, Frankel, Whitney, Landa, Lord, Orlich, King and Harwood2014; Howard et al., Reference Howard, Katsos and Gibson2019; Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Murphy et al., Reference Murphy, Burns and Kilbey2017; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). Some girls on the autism spectrum, however, had fewer friends or reciprocated friendships. Calder et al. (Reference Calder, Hill and Pellicano2013) found that the percentage of reciprocated friendship among primary school–aged girls on the autism spectrum varied considerably from 20% to 100%. Additionally, when compared to neurotypical girls, the primary school–aged girls on the autism spectrum in Dean et al. (Reference Dean, Kasari, Shih, Frankel, Whitney, Landa, Lord, Orlich, King and Harwood2014) were less likely to be nominated as a friend or as a member of a group. A notable finding of Goodall and MacKenzie (Reference Goodall and MacKenzie2019) was that the two girls on the autism spectrum did not have friends and were unsure what constituted a friend.
From the findings of Sproston et al. (Reference Sproston, Sedgewick and Crane2017) and Myles et al. (Reference Myles, Boyle and Richards2019), girls on the autism spectrum showed a preference for a small group of close friends. The girls on the autism spectrum in the study of Kuo et al. (Reference Kuo, Orsmond, Cohn and Coster2013) spent the most time with one friend of the same gender, followed by same-gender and mixed-gender friend groups. Cook et al. (Reference Cook, Ogden and Winstone2018) found that the friendships formed by girls on the autism spectrum in special and mainstream schools were frequently with other girls with special needs or on the social periphery. Choice of friends was also discussed in Tierney et al. (Reference Tierney, Burns and Kilbey2016), where participants described befriending others who help support and look after them, as seen in ‘there was something about them that was more like mothering’ (p. 78).
Activities with friends
Conversations were a common activity engaged in by girls on the autism spectrum and their friends (Kuo et al., Reference Kuo, Orsmond, Cohn and Coster2013; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). The conversation topics mentioned by the participants included fashion, popular culture, boys, gossip and private information, including secrets (Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). In Kuo et al. (Reference Kuo, Orsmond, Cohn and Coster2013), girls on the autism spectrum ranked participating in physical activities with their friends, engaging in community excursions, hanging out, visiting, and doing outdoor activities after conversation. A similarly diverse range of activities appeared in the findings of Vine Foggo and Webster (Reference Vine Foggo and Webster2017), with participants enjoying sleepovers, shopping, and going to movies and parties with their friends. When asked about their favourite activities, however, girls on the autism spectrum specified solitary activities, including art, gaming, and being with animals.
Qualities of friendship
Girls on the autism spectrum identified a range of qualities they considered important for someone to be their friend (Murphy et al., Reference Murphy, Burns and Kilbey2017; Myles et al., Reference Myles, Boyle and Richards2019; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). Companionship was a characteristic identified by the participants in Myles et al. (Reference Myles, Boyle and Richards2019) and Sedgewick et al. (Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016) and was associated with care, fun, and happiness. Acceptance in being able to be themselves and acceptance of autism were also important traits of friends (Myles et al., Reference Myles, Boyle and Richards2019; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). Trust was another common important quality and was typically linked with understanding (Murphy et al., Reference Murphy, Burns and Kilbey2017; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). Lastly, humour was considered important, with a participant stating, ‘they have to be funny. Definitely’ (Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016, p. 1302; see also Murphy et al., Reference Murphy, Burns and Kilbey2017).
Nature of friendship
Research has typically compared the nature of the friendships of girls on the autism spectrum with boys on the autism spectrum and neurotypical girls (Head et al., Reference Head, McGillivray and Stokes2014; Kuo et al., Reference Kuo, Orsmond, Cohn and Coster2013; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016). Kuo et al. (Reference Kuo, Orsmond, Cohn and Coster2013) found that both girls and boys on the autism spectrum perceived greater levels of help and closeness from female friends than male friends. Greater differences between girls and boys on the autism spectrum were found in Head et al. (Reference Head, McGillivray and Stokes2014) and Sedgewick et al. (Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016). In Head et al. (Reference Head, McGillivray and Stokes2014), girls on the autism spectrum scored higher on the Friendship Questionnaire, which measures friendship quality, understanding, and empathy, than boys on the autism spectrum but scored similarly to neurotypical boys. Friendship quality was also measured in Sedgewick et al., which showed that girls on the autism spectrum reported closer, more helpful, and more secure friendships than boys on the autism spectrum. The quality of their friendship was in fact similar to neurotypical girls regarding levels of companionship, help, closeness, and security, with the only exception being conflict.
Making and maintaining friendship
Substantial hardships were encountered by girls on the autism spectrum in their process of making and maintaining friends (Cridland et al., Reference Cridland, Jones, Caputi and Magee2014; Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Murphy et al., Reference Murphy, Burns and Kilbey2017; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). The hardships included difficulties socialising appropriately, misunderstandings in communication, struggling with disagreements and conflicts, and feeling stressed from socialising, thus withdrawing from friendships and risking the friendship weakening (Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). One girl described the communication challenges she faced being on the autism spectrum as like ‘talking to another person through an intercom machine … and somehow the wires get crossed and they get the wrong message’ (Tierney et al., Reference Tierney, Burns and Kilbey2016, p. 78). These difficulties in making and maintaining friends were related to feelings of anger, annoyance, anxiety, confusion, and exhaustion in girls on the autism spectrum (Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Murphy et al., Reference Murphy, Burns and Kilbey2017; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Tierney et al., Reference Tierney, Burns and Kilbey2016). In Hughes et al. (Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013), despite being able during pre-intervention to articulate actions to make friends and receiving an intervention using communication books, the participants still found meeting new people to be onerous. A participant in Cridland et al. (Reference Cridland, Jones, Caputi and Magee2014) also expressed the particular challenge of trying to make friends with neurotypical peers in ‘as soon as they hear that I have a disability they just won’t even listen to me’ (p. 1266).
Friendship groups
A number of studies reported on friendship groups, which were often characterised by volatility and difficulty (Cook et al., Reference Cook, Ogden and Winstone2018; Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). The size of friendship groups was varied, with some participants preferring one friend or a small group, whereas other participants mentioned being a part of a group of around 100 friends (Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016). Participants described challenges including conflict, disagreement, group communication, and social exclusion (Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). In Myles et al. (Reference Myles, Boyle and Richards2019), this social exclusion arose from feeling ignored, unwelcome, or unable to participate in the group conversation and activities. Moreover, girls on the autism spectrum described leaving groups, joining new groups, as well as being deserted by friends if they could not fit into the friendship groups (Cook et al., Reference Cook, Ogden and Winstone2018; Tierney et al., Reference Tierney, Burns and Kilbey2016).
Support with friendship
Girls on the autism spectrum received assistance from family and teachers in their friendships (Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Murphy et al., Reference Murphy, Burns and Kilbey2017; Tierney et al., Reference Tierney, Burns and Kilbey2016). Support from parents and siblings included helping to set up meetings with friends, providing practical advice, and modelling appropriate behaviour (Murphy et al., Reference Murphy, Burns and Kilbey2017; Tierney et al., Reference Tierney, Burns and Kilbey2016). In Goodall and MacKenzie (Reference Goodall and MacKenzie2019), a participant reported receiving help from their teacher to make friends, which was briefly successful, although ‘once that year ended they would stop playing with me’ (p. 508).
Peer relationships
Experiences with peers
Girls on the autism spectrum expressed various difficulties with interacting with their peers (Cook et al., Reference Cook, Ogden and Winstone2018; Dillon et al., Reference Dillon, Underwood and Freemantle2016; Myles et al., Reference Myles, Boyle and Richards2019; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Tierney et al., Reference Tierney, Burns and Kilbey2016). The participants in Dillon et al. (Reference Dillon, Underwood and Freemantle2016) perceived their peers to be distractions in the mainstream classroom. Girls on the autism spectrum in specialised settings, including pupil referral units and special schools, felt their peers misunderstood their attempts to form friendships and bullied them (Cook et al., Reference Cook, Ogden and Winstone2018; Sproston et al., Reference Sproston, Sedgewick and Crane2017). Feelings of insecurity were reported in Tierney et al. (Reference Tierney, Burns and Kilbey2016), with a participant describing that ‘I feel like a mouse and everyone else is like a giant cat …’ due to peers’ better understanding social protocols (p. 77). Similarly, the girls on the autism spectrum in Myles et al. (Reference Myles, Boyle and Richards2019) spoke about being nervous around peers owing to a limited understanding of what to do in social interactions, as well as insubstantial social skills and support.
Peer victimisation and rejection
Experiences of peer victimisation were common among girls on the autism spectrum (Cook et al., Reference Cook, Ogden and Winstone2018; Cridland et al., Reference Cridland, Jones, Caputi and Magee2014; Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Sproston et al., Reference Sproston, Sedgewick and Crane2017). The types of peer victimisation they encountered were physical, sexual, social, and verbal, including being sworn at, called names, backstabbed, punched, and having their lunch thrown around the room (Cook et al., Reference Cook, Ogden and Winstone2018; Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Sproston et al., Reference Sproston, Sedgewick and Crane2017). One girl shared how a group of boys ‘call me other names that they know I don’t like, like “Frankenstein”’ (Cook et al., Reference Cook, Ogden and Winstone2018, p. 309). Hebron and Humphrey (Reference Hebron and Humphrey2014) noted that the concept and purposes of peer victimisation were understood clearly by girls on the autism spectrum. Peer victimisation was also considered a top concern of those participants, which was echoed by the girls in Goodall and MacKenzie (Reference Goodall and MacKenzie2019). Cook et al. (Reference Cook, Ogden and Winstone2018) found that the peer victimisation faced by girls on the autism spectrum in mainstream and special settings was different, with the events occurring in special settings typically being due to other students’ special needs rather than being intentional.
Peer rejection was another significant issue faced by girls on the autism spectrum (Cook et al., Reference Cook, Ogden and Winstone2018; Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016). They shared experiences of being picked last in sport, not being invited to out-of-school activities or parties, not being listened to, underestimated, and infantilised (Cook et al., Reference Cook, Ogden and Winstone2018; Myles et al., Reference Myles, Boyle and Richards2019). A finding from Dean et al. (Reference Dean, Kasari, Shih, Frankel, Whitney, Landa, Lord, Orlich, King and Harwood2014) was that the primary school–aged boys on the autism spectrum were rejected more frequently than girls on the autism spectrum.
Fink et al. (Reference Fink, Olthof, Goossens, van der Meijden and Begeer2018) and Sproston et al. (Reference Sproston, Sedgewick and Crane2017) showed the diverse responses that girls on the autism spectrum had to bullies. The questionnaires of Fink et al. indicated the girls were more likely to provide help to victims of bullying and less likely to be passive bystanders than boys. Sproston et al. found that participants physically retaliated, withdrew, or modified their behaviour when experiencing bullying.
Classroom social connection
The social connections of girls on the autism spectrum were examined in two studies (Anderson et al., Reference Anderson, Locke, Kretzmann and Kasari2016; Calder et al., Reference Calder, Hill and Pellicano2013). None of the primary school–aged girls on the autism spectrum were socially isolated in Calder et al. (Reference Calder, Hill and Pellicano2013); however, some were more socially involved in their classrooms than others. Anderson et al. (Reference Anderson, Locke, Kretzmann and Kasari2016) found that girls on the autism spectrum in primary school formed more social connections to peers in classrooms with 21 students or more. This contrasted with boys on the autism spectrum, who had greater social connections in classrooms with 20 students or fewer.
Classroom activities
Working with peers in the classroom was generally considered by girls on the autism spectrum to be challenging. For the participants in Goodall and MacKenzie (Reference Goodall and MacKenzie2019), finding a peer to work with was embarrassing and intensified their feelings of loneliness as they typically could not find a partner. One participant described it thusly: ‘I felt so little having to stand there waiting to pair up with the teacher’ (Goodall & MacKenzie, Reference Goodall and MacKenzie2019, p. 508). For others, working with peers was difficult due to communication issues and feeling unable to cope with group tasks (Tierney et al., Reference Tierney, Burns and Kilbey2016). There was only one exception in Dillon et al. (Reference Dillon, Underwood and Freemantle2016), with a girl on the autism spectrum stating that she enjoyed working in a group if it was small, comprised of people she knew and not noisy.
Understanding self
Growing up
Difficulties in peer relationships became more prominent as girls on the autism spectrum grew up (Cook et al., Reference Cook, Ogden and Winstone2018; Tierney et al., Reference Tierney, Burns and Kilbey2016). The participants in Tierney et al. (Reference Tierney, Burns and Kilbey2016) specifically pinpointed the primary to high school transition as the threshold at which the challenges with peers became more pronounced. This was due to changes to implicit social expectations and protocol that the girls were unable to understand and conform to. Divergent interests, expectations of friendships, and maturity were also identified by girls on the autism spectrum as other differences between them and their peers and as a source of friendship difficulties (Cook et al., Reference Cook, Ogden and Winstone2018; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017).
Identity factors
Girls on the autism spectrum discussed how their autistic and gender identities impacted their relationship with peers. Participants in Goodall and MacKenzie (Reference Goodall and MacKenzie2019) and Hebron and Humphrey (Reference Hebron and Humphrey2014) felt a sense of difference from their peers due to their autistic diagnosis but did not want to be treated differently. Indeed, one girl remarked that ‘I’m happy that they’re treating me like an ordinary schoolgirl’ (Hebron & Humphrey, Reference Hebron and Humphrey2014, p. 29). Their autistic identity was framed in a more negative light in Goodall and MacKenzie, with participants regarding their autism as leading to judgement from peers and causing difficulties with social interactions. Both Cridland et al. (Reference Cridland, Jones, Caputi and Magee2014) and Tierney et al. (Reference Tierney, Burns and Kilbey2016) found that girls on the autism spectrum thought boys were more approachable than girls due to having more similar interests. Additionally, the participants in Tierney et al. felt ‘ungirly’ in comparison to other girls, leading to the perception that the female social space was inaccessible to them.
Masking
Considering the peer interaction challenges they faced, many girls on the autism spectrum compensated by modifying their behaviour (Cook et al., Reference Cook, Ogden and Winstone2018; Hebron & Humphrey, Reference Hebron and Humphrey2014; Murphy et al., Reference Murphy, Burns and Kilbey2017; Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016). These modifications included changing their personality and adopting similar actions to their peers to better fit in (Cook et al., Reference Cook, Ogden and Winstone2018; Myles et al., Reference Myles, Boyle and Richards2019). A participant in Cook et al. (Reference Cook, Ogden and Winstone2018) described her rationale as, ‘I thought if I changed to be like my other friend, they’ll listen to me, and they all did …’ (p. 310). Murphy et al. (Reference Murphy, Burns and Kilbey2017) found that the behaviour modifications could vary depending on the level of friendship, with a girl on the autism spectrum expressing that she felt more comfortable being her true self with very close friends. The participants in Tierney et al. (Reference Tierney, Burns and Kilbey2016) described using skills of empathising and observation to enhance their peer interactions and improve their ability to imitate their peers. One such observation technique was pretending to be engaged in reading while in actuality watching their peers. Girls on the autism spectrum also deliberately masked their feelings of anxiety and sadness by pretending to be neutral or happy in order to make friends or maintain their friendships.
These efforts to imitate peers and mask feelings contributed to feelings of exhaustion and depression as well as to self-harm. One girl on the autism spectrum shared that ‘I was getting to the point where I actually couldn’t control some urges that I had … I was a bit suicidal’ (Tierney et al., Reference Tierney, Burns and Kilbey2016, p. 79). A conscious concealment of emotions was also found in Hebron and Humphrey (Reference Hebron and Humphrey2014), with anger and anxiety being the feelings most often suppressed. Participants were able to find relief, however, through revealing their true feelings.
Discussion
This is the first review to provide a synthesis of qualitative, quantitative, and mixed methods research conducted between 2010 and 2020 on the peer interaction experiences of school-aged girls on the autism spectrum from their perspective. The answers to the research questions follow.
-
1. What are the peer interaction experiences of school-aged girls on the autism spectrum?
The findings highlighted that girls on the autism spectrum wanted to have friends and preferred small groups of friends (Cook et al., Reference Cook, Ogden and Winstone2018; Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016; Tierney et al., Reference Tierney, Burns and Kilbey2016). They also considered friendships to be critical in terms of emotional support and feelings of belonging (Myles et al., Reference Myles, Boyle and Richards2019; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Vine Foggo & Webster, Reference Vine Foggo and Webster2017). This is consistent with the findings of the systematic review of peer relationship experiences of adolescents on the autism spectrum conducted by Cresswell et al. (Reference Cresswell, Hinch and Cage2019), where the adolescents expressed their desire for peer relationships and belonging in society. Girls on the autism spectrum also reported feelings of loneliness due to unmet desires for friendship (Goodall & MacKenzie, Reference Goodall and MacKenzie2019; Vine Foggo & Webster, Reference Vine Foggo and Webster2017).
The findings on the importance and desire of friendships for girls on the autism spectrum challenge the longstanding assumption that people on the autism spectrum are less socially motivated than neurotypical people (Sedgewick et al., Reference Sedgewick, Hill and Pellicano2019). This underscores the importance of being informed by the under-recognised voices of girls on the autism spectrum themselves and not perpetuating incorrect stereotypes (Watkins et al., Reference Watkins, Zimmermann and Poling2014). Relatedly, considering the value girls on the autism spectrum placed on friendship, educators can support their peer interaction preferences and interests such as by ensuring a small group of their friends are present across different classes (Myles et al., Reference Myles, Boyle and Richards2019; Sproston et al., Reference Sproston, Sedgewick and Crane2017). The loneliness encountered by girls on the autism spectrum due to an unmet desire for friendship also requires attention, with research highlighting the association between loneliness and social anxiety and depression in young adults on the spectrum (Schiltz et al., Reference Schiltz, McVey, Dolan Wozniak, Haendel, Stanley, Arias, Gordon and Van Hecke2021). Social skills interventions that enable girls on the autism spectrum to further develop social skills and participate in more successful social interactions with peers could contribute to improved interactions and reduced loneliness (Schiltz et al., Reference Schiltz, McVey, Dolan Wozniak, Haendel, Stanley, Arias, Gordon and Van Hecke2021; Zeedyk et al., Reference Zeedyk, Cohen, Eisenhower and Blacher2016).
-
2. How do the peer interaction experiences of school-aged girls on the autism spectrum compare to school-aged boys on the autism spectrum?
A few notable differences between the peer interaction experiences of boys and girls on the autism spectrum became apparent through the review. Specifically, girls experienced loneliness when their desire to have friends was not met, and their levels of loneliness were also higher as compared to boys (Bossaert et al., Reference Bossaert, Colpin, Pijl and Petry2012). This aligns with the conclusions drawn in the study by Lai et al. (Reference Lai, Lombardo, Auyeung, Chakrabarti and Baron-Cohen2015), which indicated that females as opposed to males on the autism spectrum have a stronger need to interact with others and develop friendships with one or a few people. The nature of friendships experienced by girls and boys on the autism spectrum also appeared to be different. Girls had higher levels of friendship quality, understanding and empathy as compared to boys (Head et al., Reference Head, McGillivray and Stokes2014). They also reported having closer, more helpful, and secure friendships than boys on the spectrum (Sedgewick et al., Reference Sedgewick, Hill, Yates, Pickering and Pellicano2016). These differences may be due to the distinct ways that boys and girls on the spectrum interact. Kuo et al. (Reference Kuo, Orsmond, Cohn and Coster2013) found that although boys and girls on the autism spectrum had similar amounts of social interactions with peers, boys were more likely to engage in activities with friends (e.g., playing), whereas girls engaged in conversations and focused on emotional closeness. As such, girls may have greater opportunities to develop and maintain friendships characterised by empathy and of a higher quality as compared to boys (Sedgewick et al., Reference Sedgewick, Hill and Pellicano2019).
This scoping review adds to the growing body of evidence indicating gender differences in peer interaction experiences and friendship quality of girls on the autism spectrum. This highlights the need for future research to select gender-balanced samples and recognise the potential impact of gender when researching peer interaction experiences (Płatos & Pisula, Reference Płatos and Pisula2021). There is also a need for educators to be informed about the differences between the peer interaction experiences of girls and boys on the autism spectrum to provide relevant supports. Educational training and policies should account for these gender differences to be more appropriate for the needs and preferences of girls on the autism spectrum.
-
3. What peer interaction difficulties are experienced by school-aged girls on the spectrum?
The difficulties experienced by the girls in their friendships included a varying percentage of reciprocated friendship (Calder et al., Reference Calder, Hill and Pellicano2013). The most commonly experienced hardship was the process of making and maintaining friends (Cridland et al., Reference Cridland, Jones, Caputi and Magee2014; Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Murphy et al., Reference Murphy, Burns and Kilbey2017; Sproston et al., Reference Sproston, Sedgewick and Crane2017; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017), caused by difficulties with understanding the social rules of peer engagement and misunderstandings in communication. In Cresswell et al.’s (Reference Cresswell, Hinch and Cage2019) systematic review, making and maintaining friends was the most challenging for the adolescents on the spectrum, caused by not knowing how to approach others, difficulties in resolving conflict situations, or misunderstanding social conventions.
The most troubling finding of this scoping review is the continuing experience of peer victimisation and rejection, with victimisation being prevalent in its different types (physical, sexual, social, and verbal). This is consistent with Cresswell et al.’s (Reference Cresswell, Hinch and Cage2019) findings, where adolescents on the autism spectrum shared similar experiences, with some blaming themselves or their autistic traits. According to the findings from an earlier study by Sterzing et al. (Reference Sterzing, Shattuck, Narendorf, Wagner and Cooper2012), adolescents on the autism spectrum face a much higher victimisation rate (46.3%) compared to their neurotypical peers (10.6%). The finding of this scoping review also aligns with Greenlee et al.’s (Reference Greenlee, Winter and Marcovici2020) findings where adolescents on the autism spectrum reported experiencing relational victimisation, which includes more subtle forms of victimisation such as gossiping and ignoring. However, Greenlee et al.’s study indicated that relational victimisation was related more to challenges in social functioning (e.g., social cognition, awareness, communication) and restricted and repetitive behaviours in girls, but that for boys, relational victimisation was only associated with social communication impairments.
The findings of this scoping review have clear implications for supports. Given the prevalence of peer interaction challenges for girls on the autism spectrum, social skills interventions such as social skills training and peer-mediated instruction tailored for girls on the autism spectrum based on evidence-based practices should be developed (Wong et al., Reference Wong, Odom, Hume, Cox, Fettig, Kucharczyk, Brock, Plavnick, Fleury and Schultz2015). These interventions should target the specific difficulties articulated by girls, including maintaining friends into the high school years and friendship group conflicts, to improve their peer interaction experiences (Hughes et al., Reference Hughes, Bernstein, Kaplan, Reilly, Brigham, Cosgriff and Boykin2013; Myles et al., Reference Myles, Boyle and Richards2019; Tierney et al., Reference Tierney, Burns and Kilbey2016; Vine Foggo & Webster, Reference Vine Foggo and Webster2017).
Although specific anti-bullying programs for students with disability are in their infancy, it is still necessary for schools to respond to the peer victimisation and rejection that girls on the autism spectrum experience (Winchell et al., Reference Winchell, Sreckovic and Schultz2018). Educators should be conscious that girls on the autism spectrum are at increased risk of peer victimisation and rejection and foster a culture of inclusion (Yellow Ladybugs & the Department of Education and Training Victoria, 2018). Participation in anti-stigma programs can also encourage peer acceptance for girls on the autism spectrum (Ranson & Byrne, Reference Ranson and Byrne2014).
-
4. What strategies did school-aged girls on the spectrum employ when experiencing peer interaction difficulties?
The girls in the reviewed studies developed diverse approaches to overcome the challenges experienced in peer interactions. These included changing their personality to be like their peers and thus fit in (Cook et al., Reference Cook, Ogden and Winstone2018; Myles et al., Reference Myles, Boyle and Richards2019), and masking their feelings of anxiety and depression in order to make and maintain friends (Hebron & Humphrey, Reference Hebron and Humphrey2014; Tierney et al., Reference Tierney, Burns and Kilbey2016). Pretending to be like others and thus masking their autistic traits to better fit in was also one of the common coping mechanisms for women on the autism spectrum in Bargiela et al.’s study (2016), in which the women also shared detailed strategies of ‘pretending to be normal’, such as imitating other people’s accent or pretending to have an interest in television programs they did not care for (p. 3287). Similarly, participants in the Cresswell et al.’s (Reference Cresswell, Hinch and Cage2019) systematic review shared examples of masking social difficulties they encounter by pretending to be socially competent and popular.
Educators should be aware that girls on the autism spectrum use masking as a coping strategy and appreciate the significant social demands that they experience. Educators can provide appropriate support such as giving opportunities for girls to de-stress from social interactions (Vine Foggo & Webster, Reference Vine Foggo and Webster2017). This can be through providing breaks or safe spaces to be alone to recuperate from social anxiety or exertion (Goodall, Reference Goodall2018). Additionally, as masking is associated with mental health challenges such as anxiety and depression, greater research is needed to understand how to better support the mental health of girls on the autism spectrum (Mandy, Reference Mandy2019). In particular, research undertaken on adolescents who are at greater risk of mental health difficulties should be prioritised (Hebron & Humphrey, Reference Hebron and Humphrey2014).
Limitations
One limitation of this scoping review is external validity, as the included studies varied in sample size. There were no limitations posed by the authors in terms of regions included in this review; however, most studies were conducted in Western countries. There is an urgent need to explore the peer interactions of school-aged girls on the autism spectrum in non-Western countries to gain a better understanding of their experiences and any potential societal and cultural influences on them. Another limitation of this scoping review was including only peer-reviewed articles. In future reviews, including reporting from grey literature may be relevant and provide a more diverse and fuller understanding of the experiences of girls on the autism spectrum.
Conclusion
School-aged girls on the autism spectrum desired friendships but encountered challenges in making and maintaining friendships as well as victimisation and rejection. Their peer interaction experiences also appear to be divergent from boys on the autism spectrum, including the practice of masking. Further awareness of the specific needs and experiences of girls on the autism spectrum is necessary in research and schools in order to provide relevant peer interaction support critical to their development and wellbeing.
Author note
This work has been prepared from a doctoral dissertation, which will be submitted by the first author in fulfilment of the requirements for the Doctor of Philosophy at the University of New South Wales.
Funding
This work was supported by the Australian Government Research Training Program Scholarship.
