Journal of Autism and Developmental Disorders (2023) 53:3047–3064

https://doi.org/10.1007/s10803-022-05573-4

ORIGINAL PAPER

The Kind of Friend I Think I Am: Perceptions of Autistic

and Non‑autistic Young Adults
Erinn H. Finke1

Accepted: 13 April 2022 / Published online: 16 May 2022

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022

Abstract
Autistic people have different preferences for friendship than non-autistic people. The aims of the current project were to
determine how autistic people prefer to behave in their friendships and how this compares to the friendship practices reported
by non-autistic participants. Autistic (n = 102) and non-autistic (n = 107) young adults between the ages of 18 and 24 years
completed an online survey comprised of selected questions from the Friendship Questionnaire. Binary logistic regression
and multivariate general linear modeling were used to analyze and compare the responses across the groups. Results identi-
fied differences in the preferred friendship practices between the participant groups, which may further confirm the Double
Empathy Theory and provide a context for understanding the friendship normative practices of autistic people.

Keywords Young adults · Friendship · Survey research · Autism

Introduction authentic self. Belonging is considered a universal need that

extends across cultures and is included as the third rung in
Much of the published research on the topic of friendships Maslow’s (1954) hierarchy of needs, representing friend-
and autism spectrum disorders has reported on the diffi- ship, intimacy, family, and sense of connection. Belonging
culty autistic people have in making and maintaining these is experienced best through relationships (Carter, 2021), and
socially significant relationships (e.g., Bauminger & Shul- is core to the neurodiversity movement.
man, 2003; Daniel & Billingsley, 2010; Petrina et al., 2014).
The extant literature has outlined both quantitative differ- Friendship as a Targeted Outcome
ences in the numbers of friendships made and maintained
by autistic people (Mendelson et al., 2016; Petrina, et al., This suggests a need for reconsidering how the friendships
2014) as well as qualitative differences with respect to inten- of autistic individuals are conceptualized and described. It
sity and longevity compared to non-autistics (Bauminger & also indicates an opportunity to reconceptualize how inter-
Kasari, 2000; Bauminger & Schulman, 2003; Bauminger ventions are designed, as well as what should be taught (if
et al., 2008a, 2008b). Recommendations have thus typically anything), to improve friendship outcomes where desired.
been made to teach autistic people requisite social skills in Recently Donaldson et al. (2018) suggested reframing the
an attempt to improve these outcomes (e.g., DeRosier et al., discussion about the friendships of autistic people from a
2011; Laugeson et al., 2012; Solomon et al., 2004). discussion focused on social deficits to a discussion of dif-
Autistic people want to experience more than just integra- ferences in preferences and practices. In this way, the focus
tion or inclusion within communities, they want to belong would be directed at how autistic individuals want to con-
(e.g., Carter, 2016). Belonging is the feeling of security and duct their friendships and behave as friends rather than at
support when a person is accepted and valued for being their the differences in the ways autistic people think and behave
in friendships compared to their non-autistic peers. Viewing
the friendships of autistic people in this way creates a whole
* Erinn H. Finke
[email protected] new lens for examining friendships and friendship outcomes.
This lens refocuses the discussion of friendship outcomes on
1
Department of Audiology and Speech Pathology, University the norms for autistic people as a distinct group, which may
of Tennessee Health Science Center, 600 Henley Street, be similar to or different from the norms for friendships of
B034, Knoxville, TN 37996, USA

13
Vol.:(0123456789)
3048 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

non-autistic people. Instead of centering neurotypicality as Congruence in Friendship Preferences and Practices
the standard, from this perspective it is possible to provide
space for autistic individuals to express what they think, pre- Very few studies to date have collected data regarding simi-
fer, and how they want to act within their socially significant larities and differences in friendship preferences and prac-
relationships. tices of autistic people compared to those without autism
The emergence of the neurodiversity movement (Blume, (i.e., friendship congruence) from autistic people them-
1998; Singer, 2017) has changed and reshaped public and selves, though this is a stated research priority. It has been
scientific narratives about autism, has reshaped research pri- suggested that positive relationship development “may be
orities as well as the ways services are conceived and deliv- predicated on an autistic adult’s degree of fit with their part-
ered. The neurodiversity movement, and centering autistic ner’s traits and behaviors” (Morrison, et al., 2020, p. 1069).
voices, has also brought to light the ways autistic people Petrina et al. (2016) examined congruency in terms of
have attempted to conform to the cultural norms and behav- satisfaction in established friendships. These researchers
ioral expectations of a largely neurotypical world (usually to specifically investigated reciprocity (i.e., presence of friend-
the detriment of their own mental health, Sedgewick et al., ship) and mutuality (i.e., friendship quality) in friend dyads
2021). This being the case, the communication patterns of comprised of at least one autistic person. Congruence in
autistic people when they communicate with other autis- these parameters of friendship were studied using friendship
tic people and alternatively with non-autistic people is an nomination procedures to document reciprocity in friend-
emerging area of scientific inquiry (e.g., Bolis et al., 2021; ship identification. The Friendship Quality Questionnaire
Chen et al., 2021; Crompton et al., 2020; Morrison, et al., (FQQ; Parker & Asher, 1993; Webster & Carter, 2010) was
2020). General outcomes from this research suggest com- used to assess friendship mutuality. The participants in this
munication is more effective and enjoyable between matched study reported congruence within their friendships with
neurotype pairs (autistic–autistic and non-autistic–non- respect to reciprocity (42 of the 47 friendships were deter-
autistic) than between mixed neurotype pairs. Research in mined to be reciprocal) as well as interests and values. Some
this line has contributed significantly to understanding what results, however, did indicate areas of incongruence in these
has been termed the ‘double empathy problem’ theory (Mil- established friendships. The autistic participants reported
ton, 2012). The ‘double empathy problem’ theory asserts substantial differences in perception of friendship quality
that miscommunications and social dissatisfaction between when compared to the responses of their nominated friends.
autistic and non-autistic people is a result of bidirectional Conflict resolution and intimate exchange had higher con-
breakdowns rather than deficits specific to the autistic person gruence scores between the dyads than scores in the area of
alone (Milton et al., 2018). conflict and betrayal. These authors concluded the results
It is known that autistics want to have friends. In this way of the study showed the greatest difference in perceptions
they do not differ from their neurotypical peers (Bauminger related to the extent disagreements were resolved between
& Kasari, 2000). Most autistic people have reported they friends and intimate disclosure of personal information and
have at least one friend (Mendelson et al., 2016), and many feelings between the friends. Specific details about how each
have reported a desire for even more (Finke et al., 2019). group responded to the FQQ survey items was not provided,
There are several possible explanations for this lack of sat- so details regarding how the autistic participants responded
isfaction with current friendship outcomes reported for, and in terms of their preferences compared to their friends is not
by, autistics in the research literature. One may be a cultural known at a group level.
value attached to having more friends. If the social culture These reported discrepancies in congruence between
in which an autistic person lives places higher value on hav- established friends echo the findings reported by Finke et al.
ing more friendships, this could reasonably lead the autistic (2019). These researchers completed a survey of autistic and
person to report a desire for more friendships in order to be non-autistic young adults to investigate their friendship pref-
in line with that cultural value. A second explanation could erences. The results of this research project indicated differ-
be that current intervention models could be ineffective or ences in the broad perspectives and friendship preferences of
insufficient at teaching the needed knowledge and skills to the autistic and non-autistic young adults. Overall, the autis-
make and/or maintain desired friendships (Finke, 2016). A tic participants indicated a preference for less emotional and
third potential explanation is the emerging evidence of a lack physical closeness in their friendships compared to the non-
of congruence in the friendship preferences, practices, and autistic participants. Based on these results, these research-
friendship norms between autistic and non-autistic people ers suggested these discrepant patterns of preferences may
(e.g., Finke, et al., 2019), which will be the focus of the explain the higher proportion of autistic people who report
current paper. desiring and/or actually having more friends who also have
autism diagnosis (Bauminger & Kasari, 2000; Bauminger &

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3049

Schulman, 2003; Bauminger et al., 2008a; Cook et al., 2017; support the collective findings from the literature to date
Crompton et al., 2020; Locke et al., 2010). which largely describe the difficulties autistic individuals
have in making and maintaining friends in a neurotypically
Friendship Norms centered world (e.g., Petrina et al., 2014). This would also
support the ‘double empathy problem’ theory (Milton, 2012)
In making and maintaining friendships, preconceived and and first hand reports from autistic people themselves who
culturally driven expectations exist regarding how a “friend” have stated a preference for being friends and interacting
will behave (e.g., Wiseman, 1986). These expectations are with other autistics (e.g., Bauminger & Shulman, 2003;
“cognitive conceptualizations about attributes individuals Cook et al., 2017; Crompton et al., 2020).
would like their friends to possess and behaviors individuals If separate norms and expectations for friendships and
would like their friends to enact” (Hall, 2012, p. 884), and friend behaviors do exist for autistic individuals, this could
form the basis for evaluating and judging new friendships also provide the foundation for rethinking how to support
(Hall et al., 2011). Meeting or exceeding these expectations autistic social behaviors and friendship outcomes. These
has been shown to strongly predict friendship satisfaction norms could dictate the most immediate ways in which autis-
(Hall et al., 2011), while violating, or not meeting these tic people could be supported in understanding their social
expectations, has been shown to result in diminished sat- desires and preferred practices as well as the ways non-
isfaction, and even friendship termination or abandonment autistics should be educated about autistic culture without
(Clark & Ayers, 1993). In 2011, Hall completed a meta-anal- attempting to “fix” the autistic person or encouraging them
ysis of 37 studies of neurotypical friendship expectations. to blend in with the norms of the neurotypical community.
The results indicated there may be as many as 20 definable These norms could also potentially inform the ways in which
social norms for neurotypical friendships, and that these others' (i.e., non-autistic people) could be more inclusive and
norms were adhered to only marginally more rigidly by supportive of ideals, values, preferences and practices that
females than males. These data indicate, overall, that norms are different from their own.
govern all friendships, male and female alike (though the There is some data on the lack of congruence in friend-
norms can be gender specific; Felmlee et al., 2012). These ship preferences (i.e., how autistic people would prefer to
data also indicate there are many expectations placed on construct their friendships, for example, preference for phys-
being a friend, and conversely, many ways friendships can ical and emotional distance) of autistic young adults (e.g.,
fail as a result of violating these norms (whether this viola- Finke et al., 2019). However, data is also needed to under-
tion is intentional or unintentional does not seem to make stand how autistic people behave and act as friends. This
a difference). data would identify how autistics want to behave, indeed,
how they will behave in their friendships, and how to sup-
Friendship Norms for Autistic People port the formation and maintenance of the types of friend-
ships autistic people want to have.
Unlike the robust literature investigating and outlining the
norms for neurotypical friendships; friendship norms for The Current Project
autistic people have not been explicitly outlined to date.
There may be several reasons for this. One could be that The purpose of the current project was to understand how
autistic and non-autistic people have the same expectations autistic and non-autistic young adults prefer to behave in
for each other within their friendships and therefore the their friendships. This study was designed to explore the
established norms can be applied interchangeably across preferred behaviors of autistic and non-autistic young adults
both groups. Though this is highly unlikely given the recent with respect to making and keeping friendships. The data
emerging literature base supporting the ‘double empathy from the current project will provide information on the
problem’ theory (Milton, 2012) and providing consistent similarities and differences in the ways autistic and non-
evidence of autistic preference for interacting with other autistic young adults behave within their friendships and
autistic people (and vice versa for non-autistic people; Bolis how these compare to the established normative practices
et al., 2021; Chen et al., 2021; Crompton et al., 2020). for neurotypical friendships that have been reported in the
A second explanation may be the prevalence of medical research literature. The questions asked in the current project
model of disability approaches within the field of autism. were stated as hypothetical choices for actions within the
The overarching goal of medical model approaches is to context of interacting with a friend (e.g., “When I talk with
“treat” autism through the provision of intervention that my friends on the phone, it is usually to make arrangements
teaches autistic people how to behave more like their neuro- rather than to chat”, or “When I talk with my friends on
typical peers (Kapp et al., 2012). The existence of separate the phone, it is usually to chat rather than to make arrange-
and divergent norms for friendship for autistic people would ments”; Friendship Questionnaire (FQ); Baron-Cohen &

13
3050 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Wheelwright, 2004). The questions were not about specific friendships of adults with Asperger Syndrome. The FQ, as
friends or specific friendships, but rather centered on general originally designed and validated, is a self-report survey
normative practices within friendships based on a neurotypi- that uses forced-choice, multiple choice, and Likert scale
cal perspective of friendship (Hall et al., 2011; Wiseman, questions to quantify the gender differences in how relation-
1986). Given this, the aims of the current project were first to ships are conducted by autistic people. The FQ items assess
determine the preferred friendship behaviors of autistic and the quality of the respondents’ friendships, the respond-
non-autistic young adults with respect to general friendship ents’ level of investment in their friendships, as well as the
contexts an actions and second to determine the similarities enjoyment derived from social interaction with other people
and differences between these groups to identify congruence through a series of forced choice questions based on estab-
or incongruence in preferred friendship behaviors based on lished friendship norms for neurotypical people (Felmlee
neurotypical normative friendship practices. et al., 2012; Hall, 2011).
To investigate the aims of the current project, eleven
questions were adopted from the original FQ. The questions
Method included are items that explore preferred friendship behav-
iors using friendship norm referenced statements. The intent
Research Design was to determine how autistic young adults prefer to act
within their friendships and to compare that to the preferred
Data for this study were collected as part of a larger study friendship behaviors of non-autistic young adults. Hyman
investigating the friendships and romantic relationships of et al. (2006) reported several advantages to using “recycled”
autistic young adults. Approval was obtained through a uni- questions from a validated instrument. The primary advan-
versity Institutional Review Board (IRB) prior to initiation tage includes knowing the survey items accurately measure
of any research activities. Consistent with the IRB approved the construct of interest. This can result in obtaining data
procedures, implied informed consent was obtained from of higher quality over creating new questions that have not
all of the autistic and non-autistic young adult participants previously been tested or examined for validity (Hyman
prior to their participation in the current project. Participa- et al., 2006). For this reason, questions from a previously
tion consisted of completion of a self-administered Qual- validated survey instrument designed to assess friendship
trics™ online survey about preferred friendship behaviors relative to established normative friendship practices were
and romantic relationships. The full survey included ques- selected for use in the current project, despite being used to
tions about the participants preferences and perspectives on answer a different question than originally intended. Because
friendship generally, their knowledge of the similarities and the FQ was not used in its entirety for the current project an
differences between friends and acquaintances, as well as FQ score was not calculated. See Appendix 1 for specific
their perspectives on their preferences and behaviors within survey question items.
romantic relationships. Thirty autistic and non-autistic young adults (15 with and
To allow the researchers to collect data from a large and 15 without) piloted the survey questions used in the current
geographically dispersed sample, participants were recruited project prior to distribution of the survey for data collection.
through the Interactive Autism Network (van Selm & Pilot participants were asked to comment on the clarity and
Jankowski, 2006) database. A survey research method was interpretability of the survey items as well as the length of
selected to answer the research question of interest. Online the survey. The autistic pilot participants were also asked
surveys have been demonstrated to be an effective method about the demographic sections of the study and whether
for collecting data from autistic people about a variety of additional information should be asked or if the order of
topics including their healthcare and educational preferences the questions should be modified. The survey was revised
and practices (Nicolaidis et al., 2013; Parsons, 2015). Get- based on feedback from the pilot participants. Specific
ting a first-person perspective from participants while also changes were recommended about including instructions at
limiting the interviewer bias and the data entry error rates the beginning of each subsection of the survey as well as
of other research approaches are other established benefits modifying the grouping of questions and the skip patterns
of survey research methodology (Rea & Parker, 2005; van (i.e., a skip pattern was used to advance participants through
Selm & Jankowski, 2006). the demographic survey questions based on their responses,
so every participant did not respond to every demographic
Survey Development question). None of the actual friendship survey questions
adopted from the FQ required revision based on feedback
The survey questions for the current project were adopted from the pilot participants.
from the Friendship Questionnaire (FQ) developed by
Baron-Cohen and Wheelwright (2004) to assess the

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3051

Inclusion Criteria posted the approved recruitment message herself. It has been
established that using social media sites for recruiting young
Inclusion criteria were established to locate respondents who adults to participate in research activities is an appropriate
had characteristics and experiences relevant to the research strategy as many young adults report using these networking
question (Creswell, 1998). To participate in the study, an sites (Smith & Anderson, 2018). Second, some individuals
individual needed to: (1) be at least 18 years old; (2) have or groups further shared the recruitment message with other
either a confirmed diagnosis of ASD or report no history of groups or their “friends” or “followers” through the use of
an ASD diagnosis; (3) have access to the internet; and (4) information diffusion functions on social media platforms
provide implied informed consent to participate in the study. (e.g., “share” on Facebook; “retweet” on Twitter).
Participants who did not meet the inclusion criteria were not
eligible to complete the survey. Responses were collected Participants
from young adults between the ages of 18 and 24 with and
without a diagnosis of ASD for the purpose of being able to A total of 102 autistic young adults and 107 non-autistic
compare across groups. young adults between the ages of 18 and 24 completed
the survey questions included in the analysis for the cur-
Recruitment of Autistic Participants rent project. Autistic respondents had a mean age of
21.74 years while the non-autistic respondents had a mean
Autistics were recruited as participants for this study through age of 22.20 years. The range in ages represented across both
assistance from the Interactive Autism Network (IAN) prior groups was 18 to 24 years. The majority of the non-autistic
to the closing of the network in June 2019. The Interac- participants were female (72%; n = 77), non-Hispanic/Latino
tive Autism Network was a research network established in (96.3%; n = 103), White (87.9%; n = 94), and were, at the
2006 which was located at the Kennedy Krieger Institute. time of survey completion, enrolled in some level of college
The Interactive Autism Network was sponsored by Ken- education (65.4%; n = 70).
nedy Krieger Institute and the Simons Foundation in order The majority of the autistic participants were male
to better facilitate research regarding autism, its causes, and (71.6%; n = 73), non-Hispanic/Latino (94.1%; n = 96), White
advancement of treatment. The goal of IAN was to give (85.3%; n = 87), and were, at the time of survey comple-
autistic individuals and their families the opportunity to tion, enrolled in some level of high school or college edu-
share their thoughts and opinions with researchers. There cation (96.1%; n = 98). Specific ASD diagnoses reported
were over 17,000 registered autistic participants in IAN, by the autistic participants recruited from the IAN data-
which provided a large pool of participants to sample for base included Autism (34.3%; n = 35), Asperger Syndrome
research purposes. This method of recruitment provided the (52.9%; n = 54) and Pervasive Developmental Disorder-Not
researchers the ability to have documented confirmation of Otherwise Specified (12.7%; n = 13). For additional demo-
the participants’ ASD diagnosis, monitor the progress of graphic information and comparison statistics for both par-
participation, screen for eligibility, and offer incentives for ticipant groups, see Table 1.
participation. For this project, each individual was provided
access to the URL with an explanation of the importance Procedure
and intention of the research study. Per IAN protocol, the
participants who completed the survey received a ten-dollar After responding to a recruitment message and accessing the
gift card incentive. online survey, the participants were provided with informa-
tion about the purpose of the study, the time requirement
Recruitment of Non‑autistic Participants for participation and a statement of their rights as a research
participant. Participants then responded to demographic
Posting information about the survey on social media questions. Two questions in the demographic information
platforms was the primary method used to recruit non- section were screening questions. First, if a respondent indi-
autistic individuals. To recruit participants in this way, an cated they were not over the age of 18, the survey auto-
approved recruitment message (along with the link to the matically skipped to the end and thanked the respondent
online survey) was posted on various social media websites for their participation. Second, if the respondent indicated
(e.g., Facebook and Twitter). Most of the posts on Face- they did not have a diagnosis of ASD then they skipped the
book were directed at groups that attract young adults in demographic questions associated with being autistic. The
the desired demographic. The information about the study second section of the survey included the questions adopted
on these pages was generally posted by the moderators of for the purpose of answering the research questions. The
these groups after obtaining their approval. If the group respondents were reminded they could exit and return to
was public and/or did not have a moderator, the researcher the survey provided they used the same computer and same

13
3052 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Table 1  Participant group demographic information

Variable Autistic participants Non-Autistic participants χ2
(n = 102) (n = 107)

Age 18–20 (n = 50) 18–20 (n = 18)

21–22 (n = 26) 21–22 (n = 34)
23–24 (n = 26) 23–24 (n = 54)
χ2 = (10, 209) = 33.76, < .001
Gender Male (n = 73) Male (n = 30)
Female (n = 29) Female (n = 77)
χ2 = (1, 209) = 39.59, < .001
Education Attained Highschool (n = 64) Highschool (n = 34)
Associates (n = 11) Associates (n = 4)
4-year (n = 23) 4-year (n = 40)
Graduate (n = 4) Graduate (n = 29)
χ2 = (4, 209) = 36.18, < .001
Ethnicity Not Hispanic/Latino (n = 96) Not Hispanic/Latino (n = 103)
Hispanic/Latino (n = 5) Hispanic/Latino (n = 3)
χ2 = (1, 207) = .626, .43
Race White (n = 87) White (n = 94)
Black/AA (n = 4) Black/AA (n = 4)
Asian (n = 2) Asian (n = 7)
Native Hawaiian (n = 0) Native Hawaiian (n = 0)
American Indian (n = 3) American Indian (n = 0)
Asian and White (n = 2) Asian and White (n = 1)
American Indian and White (n = 2) American Indian and White (n = 0)
Asian and Native Hawaiian (n = 1) Asian and Native Hawaiian (n = 0)
American Indian, Black, White (n = 1) American Indian, Black, White (n = 0)
χ2 = (7, 208) = 10.31, .17

web browser to access the survey each time. Once the par- Results
ticipant submitted their responses to the survey items they
were thanked for their participation and could no longer gain Table 2 presents the results of the binary logistic regression
access to the survey. analyses for the preferred friendship behaviors examined in
the current investigation. Individual results for each variable
Data Analysis examined will be reported below. The complete results for
each regression analysis for each variable are reported in
After all data were collected, the data were cleaned and Appendix 2.
checked for outliers. Descriptive statistics were used to
describe responses to the various demographic questions. I Would Rather Talk on the Phone with a Friend Just
Binary logistic regression was used to determine the odds to Chat or to Make Arrangements
ratios of autistic individuals reporting similar responses as
the non-autistic participants for the binary forced choice Reported autism status and gender were entered into the
questions. Gender and ASD status were included in all regression model as predictors of reported preferred friend-
logistic regression models to account for any differences as ship behaviors. When examining responses to the survey
a result of these demographic factors across and within the item “I would rather talk on the phone with a friend just
groups. The interaction variable between ASD status and to chat” or “I would rather talk on the phone with a friend
Gender was also included in all of the models. Multivariate to make arrangements,” the overall model was statistically
general linear modeling was used to analyze and compare significant, X(3) = 19.41, p < 0.001. The overall sensitivity
the responses to the scale items across the groups (autistic of the model for accuracy in classification was 76.2% and
and non-autistic). Data analysis for this study was completed 65.3% of the total cases were categorized correctly. Overall,
using SPSS version 28, a statistical analysis program. 202 participants responded to this item (95 autistics; 107

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3053

Table 2  Regression results summaries

Preferred friendship behavior Intercept ASD Gender ASD x Gender
Sig Exp(B) Sig Exp(B) 95% CI Sig Exp(B) 95% CI Sig Exp(B) 95% CI

Chat (0) v Arrange (1) .57 1.08 .007 3.54 1.42–8.83 .002 4.16 1.70–10.20 .03 .24 .07–.88
Talk (0) v Activity (1) .09 1.27 < .001 4.75 1.91–11.82 .03 2.68 1.11–6.46 .96 1.04 .27–3.93
Fun (1) v Support (0) .02 .72 .008 3.71 1.42–9.71 .004 4.02 1.55–10.38 .75 .81 .22–2.98
Feelings (0) v .78 .96 .29 1.61 .67–1.93 .009 3.20 1.33–7.70 .35 .55 .16–1.93
Solutions (1)
Wait (1) v Contact (0) .10 1.26 .004 4.03 1.58–10.26 .43 .69 .28–1.72 .31 2.01 .52–7.82
Blunt (1) v Gentle (0) .008 .68 .95 .97 .36–2.63 .095 2.16 .87–5.31 .13 2.79 .73–10.64
Personality (same (0) v diff (1)) < .001 .27 .08 2.74 .89–8.43 .19 2.16 .68–6.85 .64 .70 .15–3.19
Interests (same (0) v diff (1)) < .001 .18 .94 1.07 .20–5.83 .54 1.60 .36–7.16 .25 3.47 .41–29.43
Meet up with friends < .001 .49 < .001 8.96 3.26–24.68 .53 .64 .16–2.53 .97 1.04 .20–5.33
(weekly (0) v monthly or less (1))
Online only friends < .001 .24 .003 5.26 1.77–15.67 .69 .71 .10–3.65 .908 1.12 .17–7.27
(zero (0) v some (1))

non-autistics; see Table 3). Results of the binary logistic meeting friends to chat compared to people without an ASD
regression indicated autistic individuals were 3.5 times (OR diagnosis, X(1) = 11.23, p < 0.001. Non-autistic people most
3.54, p.007) more likely to report they would rather talk frequently reported they would prefer to talk on the phone
on the phone with a friend to make arrangements (63.2%; with a friend just to chat (57.9%; n = 62/107).
n = 60/95) compared to people without an ASD diagnosis,
X(1) = 7.32, p < 0.007 who most frequently reported they My Friends Value Me as a Support or as Someone
would prefer to talk on the phone with a friend just to chat to Have Fun With
(57.9%; n = 62/107).
When examining responses to the survey item “My friends
I Would Rather Meet Up with a Friend to Chat value me as someone who us a support to them” or “My
or for a Specific Activity friends value me as someone to have fun with,” the overall
model was statistically significant, X(3) = 44.89, p < 0.001.
When examining responses to the survey item “I prefer The overall sensitivity of the model for accuracy in classifi-
meeting friends for a specific activity” or “I prefer meeting cation was 56.5% and 71.4% of the total cases were catego-
friends to chat,” the overall model was statistically signifi- rized correctly. Overall, 203 participants responded to this
cant, X(3) = 49.77, p < 0.001. The overall sensitivity of the item (98 autistics; 105 non-autistics; see Table 3). Results
model for accuracy in classification was 80.7% and 72.1% of the binary logistic regression indicated autistic individu-
of the total cases were categorized correctly. Overall, 204 als were 3.7 times (OR 3.71, p 0.008) more likely to report
participants responded to this item (98 autistics; 106 non- their friends value them as someone to have fun with (61.2%;
autistics; see Table 3). Results of the binary logistic regres- n = 60/98) compared to people without an ASD diagnosis,
sion indicated autistic individuals were 4.8 times (OR 4.75, X(1) = 7.11, p < 0.008 who most frequently reported their
p. < 001) more likely to report they would prefer to meet up friends value them as someone to support them (76.2%;
with a friend for a specific activity (78.6%; n = 77/98) over n = 80/105).

Table 3  Summary of participant responses by group and gender

Gender ASD Chat/arrange Talk/activity Fun/support Solu- Wait/contact Blunt/gentle Personality Interests Meet up Online only
tions/
feelings

Female No 77 77 75 77 76 77 77 77 68 77
Yes 27 29 29 28 29 29 29 29 28 29
Male No 30 29 30 30 29 29 30 30 30 30
Yes 68 69 69 63 65 67 71 72 65 71

13
3054 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

I Am Better at Discussing Feelings or Coming Up If I Had Something to Say to a Friend, I Would Say It
with Practical Solutions Directly or I Would Broach It Gently

The binary logistic regression analysis results examining The analysis examining the survey item if I had something to
whether autistic and non-autistic participants viewed them- say to a friend, I would broach the subject gently or I would
selves as better at discussing feelings or better at coming just come out and say it indicated the overall model was
up with practical solutions indicated the overall model was statistically significant, X(3) = 29.22, p < 0.001. The overall
statistically significant, X(3) = 11.90, p.008. The overall sensitivity of the model for accuracy in classification was
sensitivity of the model for accuracy in classification was 53.7% and 69.8% of the total cases were categorized cor-
58.8% and 61.6% of the total cases were categorized cor- rectly. Overall, 202 participants responded to this item (96
rectly. Overall, 198 participants responded to this item (91 autistics; 106 non-autistics; see Table 3). Results indicated
autistics; 107 non-autistics; see Table 3). There was no neither ASD (p .95; OR 1.03); gender (p < 0.095; OR 0.46);
difference based on ASD diagnosis. Males, with and with- nor the ASDxGender interaction (p < 0.13; OR 0.36) were
out an ASD diagnosis (61.3%; n = 57/93), were 3.2 times significant. Both autistic (75.9%; n = 22/29) and non-autistic
(OR = 3.20, p.009) more likely to say they are better at (75.3%; n = 58/77) females reported they would broach the
coming up with practical solutions than female participants subject gently with a friend. This was the same for the major-
(38.1%; n = 40/105). ity of non-autistic males (58.6%; n = 17/29). The majority
of autistic males, however, reported they would just come
I Would Contact My Friend to Discuss a Problem or I out and say what they needed to say to a friend (65.7%;
Would Wait for My Friend to Contact Me n = 44/67).

When examining the preferences of the autistic and non- Reactions to Fallouts with Friends
autistic respondents with respect to whether they would
contact a friend to discuss a problem or whether they would The next question on the survey was a Likert scale ques-
wait for the friend to contact them first, the data indicated tion that asked the respondents to report how they would
the overall model was statistically significant, X(3) = 33.43, react to a fallout with a friend. The four response options
p < 0.001. The overall sensitivity of the model for accuracy for all of the questions were, if there were a fallout in my
in classification was 64.9% and 69.3% of the total cases were friendship, I: (1) would do whatever it takes to repair the
categorized correctly. Overall, 199 participants responded to relationship; (2) would make the first move, as long as they
this item (94 autistics; 105 non-autistics; see Table 3). Over- reciprocated; (3) would sort it out, if they make the first
all, there were significant differences based on ASD status move; (4) would not be able to be close friends anymore.
(OR = 4.03, p.004). Autistic participants were 4 times more Cross tabulations and Pearson Chi-Square were used to com-
likely to report they would wait for their friend to contact pare the responses of the respondents. See Tables 4 and 5 for
them (76.6%; n = 72) compared to non-autistic participants, the results of these comparisons. Overall, the data indicated
X(1) = 8.51, p.004 who more frequently reports they would all of the comparisons were significant. The total Pearson
contact their friend to discuss the problem (62.9; n = 66). Chi-Square p < 0.000. A discrepancy between the groups
occurred in response category 4 (I would not be able to be
close friends anymore). None of the non-autistic partici-
pants, male or female, selected this response option while
17.3% of the autistic participants (13.8% of females, 18.8%

Table 4  Reactions to fallouts with friends

Fallout variables Female Male Total

If there were a fallout in my friendship, I am willing to make the first move ASD = 37.9% (11) ASD = 29.0% (20) ASD = 31.6% (31)
NT = 66.2% (51) NT = 60.0% (18) NT = 64.5% (69)
If there were a fallout in my friendship, I am willing to sort it out, if they make the ASD = 20.7% (6) ASD = 23.2% (16) ASD = 22.4% (22)
first move NT = 13.0% (10) NT = 26.7% (8) NT = 16.8% (18)
If there were a fallout in my friendship, I would not be able to be close friends ASD = 13.8% (4) ASD = 18.8% (13) ASD = 17.3% (17)
anymore NT = 1.3% (1) NT = 0.0% (0) NT = 0.9% (1)
If there were a fallout in my friendship, I would do whatever it takes to repair the ASD = 27.6% (8) ASD = 29.0% (20) ASD = 28.6% (28)
relationship NT = 19.5% (15) NT = 13.3% (4) NT = 17.8% (19)

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3055

Table 5  Group comparisons for reactions to fallouts with friends Table 7  Group comparisons for personality and friends
Pearson Chi-Square Pearson Chi-Square
Value df Significance Value df Significance

Female 11.323 3 .008** Female 6.715 3 .069

Male 13.109 3 .004** Male 3.822 3 .288
Total 30.449 3 .000*** Total 10.073 3 .015*

Note. p < .05*, p < .01**, p < .001*** p < .05*, p < .01**, p < .001***

of males) selected this response option. There was also a In Terms of Interests, How Similar to Your Friends Do
discrepancy in the participants’ reported willingness to make You Tend To Be?
the first move to repair the relationship with a friend after a
fallout between the autistic and non-autistic groups. 64.5% The next question on the survey was a Likert scale question
of the non-autistic participants selected this response, while that asked the respondents to report how similar or differ-
only 31.6% of the autistic respondents indicated this would ent they think they are compared to their friends in terms
be their path forward with a friend. of their interests. The four response options for all of the
questions were, (1) very similar, (2) quite similar, (3) not
In Terms of Personality, How Similar to Your Friends very similar, and (4) very dissimilar. Cross tabulations and
Do You Tend To Be? Pearson Chi-Square were used to compare the responses of
the respondents. See Tables 8 and 9 for the results of these
The next question on the survey was a Likert scale question comparisons. Overall, the data indicated none of the com-
that asked the respondents to report how similar or different parisons were significant (based on ASD status or gender),
they think they are compared to their friends in terms of per- but the overall Pearson Chi-Square was significant p < 0.007.
sonality. The four response options for all of the questions Based on these results, the data were collapsed into two
were, (1) very similar, (2) quite similar, (3) not very similar, categories representing “similar” and “dissimilar” and a
and (4) very dissimilar. Cross tabulations and Pearson Chi- binary logistic regression was completed to further examine
Square were used to compare the responses of the respond- the responses of the participants (see Table 2). The results
ents. See Tables 6 and 7 for the results of these comparisons. of the binary logistic regression also showed no differences
Overall, the data indicated none of the comparisons were between the groups based on ASD status or gender. The
significant (based on ASD status or gender), but the overall ASDxGender interaction was also not significant. Overall,
Pearson Chi-Square was significant p < 0.015. the majority of the participants across all groups reported
Based on these results, the data were collapsed into two their interests are similar to those of their friends.
categories representing “similar” and “dissimilar” and a
binary logistic regression was used to examine the responses How Often Do You Meet Up with Your Friends Who
of the participants (see Table 2). The results of the binary Live Near You?
logistic regression also showed no differences between the
groups based on ASD status or gender. The ASDxGender The binary logistic regression analysis examining the pref-
interaction was also not significant. Overall, the majority of erences of the autistic and non-autistic respondents with
the participants across all groups reported their personalities respect to how often they meet up with their friends who
are similar to those of their friends. live near them indicated the overall model was statistically

Table 6  In terms of personality, Personality variables Female Male Total

how similar to your friends do
you tend to be? Very similar ASD = 20.7% (6) ASD = 16.9% (12) ASD = 18.0% (18)
NT = 11.7% (9) NT = 30.0% (9) NT = 16.8% (18)
Quite similar ASD = 55.2% (16) ASD = 50.7% (36) ASD = 52.0% (52)
NT = 77.9% (60) NT = 50.0% (15) NT = 70.1% (75)
Not very similar ASD = 24.1% (7) ASD = 26.8% (19) ASD = 26.0% (26)
NT = 9.1% (7) NT = 20.0% (6) NT = 12.1% (13)
Very dissimilar ASD = 0.0% (0) ASD = 5.6% (4) ASD = 4.0% (4)
NT = 1.3% (1) NT = 0.0% (0) NT = 0.9% (1)

13
3056 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Table 8  In terms of interests, Personality variables Female Male Total

how similar to your friends do
you tend to be? Very similar ASD = 31.0% (9) ASD = 26.4% (19) ASD = 27.7% (28)
NT = 19.5% (15) NT = 36.7% (11) NT = 24.3% (26)
Quite similar ASD = 62.1% (18) ASD = 44.4% (32) ASD = 49.5% (50)
NT = 74.0% (57) NT = 53.3% (16) NT = 68.2% (73)
Not very similar ASD = 6.9% (2) ASD = 22.2% (16) ASD = 17.8% (18)
NT = 6.5% (5) NT = 6.7% (2) NT = 6.5% (7)
Very dissimilar ASD = 0.0% (0) ASD = 6.9% (5) ASD = 5.0% (5)
NT = 0.0% (0) NT = 3.3% (1) NT = 0.9% (1)

significant, X(3) = 38.31, p < 0.001. The overall sensitivity Discussion

of the model for accuracy in classification was 79.4% and
70.7% of the total cases were categorized correctly. Over- The aim of the current investigation was to identify the simi-
all, 191 participants responded to this item (93 autistics; larities and differences in the preferred friendship behav-
98 non-autistics; see Table 3). Overall, autistic participants iors of autistic and non-autistic young adults. The results
were nearly nine times more likely to report meeting up with of the current project revealed some interesting differences
their friends who live near them monthly or less (53.8%; based on gender that were not anticipated and that should
n = 50/93) compared to the non-autistic participants (13.3%; be acknowledged. For many of the questions, the data indi-
n = 13/98) who were more likely to report meeting up with cated the non-autistic male participants responded similarly
their friends who live near them weekly (86.7%; n = 85/98). to autistic males and females. There are a couple potential
explanations for this pattern of results. First, and probably
How Many Close Friends Do You Have That You’ve most likely, is that the sample of non-autistic males in the
Never Met “In Real Life” or Face‑to‑face? current study was too small, which is a limitation that should
be considered in the interpretation of the results presented
When examining the responses from the participants regard- below. Alternatively, these responses could represent an
ing the number of friends they have that are “online only” emerging new pattern in the preferred friendship behaviors
friends. Data indicated the overall model was statistically of non-autistic males that should be studied more intently in
significant, X(3) = 18.02, p < 0.001. The overall sensitivity future investigations.
of the model for accuracy in classification was 100% and From the outset of the current project, it was hypothesized
80.7% of the total cases were categorized correctly. Overall, that, like the lack of congruence observed in the friendship
207 participants responded to this item (100 autistics; 107 preferences of autistic and non-autistic young adults (Finke
non-autistics; see Table 3). Overall, there were significant et al., 2019), there would also be a lack of congruence in the
differences based on ASD status (OR 5.26, p.003). Autistic ways autistic and non-autistic young adults act and behave
participants were approximately five times more likely to as friends. The findings from the current study largely sup-
report having friends they had never met “in real life” (31%; ported this hypothesis that autistic and non-autistic people
n = 31/100) compared to the non-autistic participants (8.4%; behave differently as friends (with some exceptions when
n = 9/107). considering the responses provided by non-autistic males).
These patterns of preferred friendship behavior, those iden-
tified in the current study and by researchers in previously
published investigations of autistic friendships, are impor-
tant to identifying and describing the culture of autistic
friendship and for establishing preferred patterns regarding
Table 9  Group comparisons for interests and friends how to be a friend.

Pearson Chi-Square
Autistic Friendships
value df Significance
The data are mounting to suggest autistic individuals have
Female 1.673 2 .455
preferences about how they want to act in their socially sig-
Male 4.489 3 .237
nificant relationships and that these preferences may differ
Total 11.718 3 .007**
from the preferences and actions of some (if not many or
p < .05*, p < .01**, p < .001*** most) non-autistic people (Brownlow et al., 2015; Calder

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3057

et al., 2013; Dean et al., 2014; Finke et al., 2019; Long et al., indicate fairly robustly that autistics spend the majority
2018). It therefore seems warranted to center the discussion of their time with friends doing an activity (e.g., playing
of friendship outcomes of autistic people as a distinct group. videogames, board games, playing on the computer) and
It now seems the evidence is becoming robust enough (based less time purely hanging out (Bauminger & Kasari, 2000;
on the data from the current project as well as the findings Bauminger & Shulman, 2003; Carrington et al., 2003; Dean
from previously published studies in this area, see above) to et al., 2017; Kuo et al., 2013; Petrina, et al., 2014). Kuo and
proceed from the perspective that autistic people largely do colleagues (2014) also reported an autistic preference for
not think or behave like non-autistic people. being with their friends in person rather than talking with
them on the phone. Finally, Daniel and Billingsley (2010)
Congruence in Broad Perspectives and Friendship reported autistic boys preferred making new friends within
Practices the context of activities and an autistic preference for spend-
ing time with established friends doing activities related to
Though it is well established that congruence between shared interests together.
friends is important to friendship intensity and longev- The finding from the current survey indicating autis-
ity (Bagwell & Schmidt, 2011), congruence of preferred tic participants believed their friends value them more as
friendship behaviors likely does not need to be tied to a someone to have fun with rather than as a support system,
specific cultural norm (Brownlow et al., 2015; Donaldson and is consistent with findings from previously published
et al., 2018; Finke et al., 2019). In examining the findings studies indicating autistic young adults preferred friend-
from the current survey in relation to findings about autistic ships that were less emotionally intense and required less
friendships that have been published in the research litera- physical proximity than their non-autistic peers (Finke et al.,
ture to date there is consistency in how autistic people have 2019). Considering these findings in relation to the broader
described and defined their friendships. In the past these literature on autistic friendships there is, again, consist-
descriptions have been compared to neurotypical norms for ency in how autistic people have discussed their friendship
friendship preferences and practices and have been found preferences and practices. The difference, again, is the lens
deficient and in need of remediation. Looking at these in which these descriptions are viewed. For years autistics
findings through the lens of the social model of disability have been telling researchers they have a preference for less
affords, however, an alternative perspective to these patterns closeness (both emotional and physical) with their friends
(Donaldson et al., 2018). This perspective is that the friend- (Calder et al., 2013; Dean, et al., 2017; Kuo et al., 2013). In
ship preferences and practices of autistic people are simply 2003 Calder and colleagues reported autistic children want
different from those of many non-autistic people. to have friends but describe their friends as companions and
As with previously published studies (e.g., Finke et al., don’t want to constantly be with other people. Dean et al.
2019; Petrina et al., 2014) results of the current project (2017) also reported autistic boys prefer having more time
identified differences in the broad perspectives and friend- alone when compared to autistic girls and their non-autistic
ship practices of autistic and non-autistic young adults. peers. These researchers also reported autistic girls to be
These findings may assist with interpreting the findings more talkative than autistic boys and to have a higher pref-
reported previously in the research literature on the topic of erence for talking with friends as an activity than autistic
the friendships of autistic people (e.g., Calder et al., 2013; boys. Kuo et al. (2013) also reported autistic girls to be more
Daniel & Billingsley, 2010; Dean et al., 2014, 2017; Kuo likely to engage in “social” activities with their friends than
et al., 2013). The results across this published literature have autistic boys.
largely been consistent across several decades of inquiry on The results of the current study indicated autistics would
the topic of autistic friendship preferences and practices. be less likely to initiate contact with a friend to discuss a
The results of the current project indicated autistic young problem or resolve a disagreement. Both autistic males
adults would rather talk on the phone to make arrangements and autistic females who responded to the current survey
than simply to chat. Of the current sample, nearly 60% of reported a reluctance to initiate contact with friends in the
the autistic respondents (58.8%; n = 60/102) would prefer midst of a disagreement (78.5%; n = 51/65 of autistic males;
to talk on the phone to make plans. The results of the cur- 72.4%; n = 21/29 of autistic females). This, again, is consist-
rent survey similarly showed that the autistic participants ent with the ways autistic people have reported behaving
preferred meeting up with their friends to participate in a in friendships in the published literature. This friendship
specific activity (75.5%; n = 77/102) over meeting up with practice has been reported as a preference for avoiding con-
them to hang out or chat. These findings are consistent with flict with others, especially friends (Carrington et al., 2003;
the previously published research literature (though most of Daniel & Billingsley, 2010). Petrina et al. (2016) reported
the studies that have been published to date included par- incongruence in the established friendships of autistics with
ticipants who were younger than the current sample) which respect to how, when and if disagreements were resolved.

13
3058 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

This reluctance to engage in conflict or conflict resolution Implications

could also be related to the finding reported in the litera-
ture that autistic people have friendships that are shorter in The results of the current study indicate autistic young adults
duration compared to their non-autistic peers (Bauminger prefer to behave in friendships in ways that are different from
& Shulman, 2003). non-autistic young adults in some important ways. Identify-
It is further interesting to consider the different ways ing differences is a fairly common outcome when comparing
autistic and non-autistic young adults would react to a fall- data collected from autistic and non-autistic participants.
out with a friend. Nearly 20% of the autistic participants Differences between autistics and non-autistics have been
reported they would no longer be able to be close friends reported for as long as these comparisons have been made.
with someone with whom they had a fallout or disagree- The interesting implications of the current research find-
ment. This is noteworthy because not one single non-autistic ings emerge from the consideration of what these data (from
participant selected this response option. In addition, non- this project and collectively across the body of literature
autistics were more than twice as likely to indicate they examining the friendships of autistic people) mean regarding
would make the first move to repair a friendship than autistic how we move forward and think about autistic friendships
participants, which is interesting, but should not be surpris- from this point forward. The data are clear and consistent.
ing. Again, autistics have been reporting their preference to Autistic people have been explaining how they have friend-
avoid conflict in socially significant relationships for many ships and what friendship means to them for many years.
years (e.g., Carrington et al., 2003; Daniel & Bilingsley, It is time these perspectives and experiences are regarded
2010). Previously published research has also indicated as valid even though they are outside of the neurotypical
that autistics, and particularly autistic girls, actively avoid standard norms. It is time to reframe the interpretation of
conflict with their friends in order to maintain friendships this information from seeing differences as deficits to seeing
(Vine Foggo & Webster, 2017). It has also been reported these differences as preferences.
that this desire to avoid conflict and maintain social relation- This is a new way to think about, talk about, and design
ships with friends has led autistic girls to mask their autism interventions related to friendship outcomes for autistic chil-
symptoms to better fit in with their non-autistic peers (Cook dren and adults. Using this perspective, the focus becomes
et al., 2017). more about what the autistic person wants for themselves
Finally, the data from the current project suggest both and less about helping the autistic person adopt the non-
autistic and non-autistic participants maintain friendships autistic norms and practices for friendships. It is less about
with people who have personalities and interests that are teaching autistic people to behave more neurotypically.
similar to their own, which is consistent with what has been Research has clearly shown that autistic people want to
reported previously about autistic friendships (Carrington have friends. Most autistics have reported having at least one
et al., 2003; Daniel & Billingsley, 2010; Kuo et al., 2013; friend (Mendelson et al., 2016), and many have reported a
Petrina et al., 2016). The non-autistic participants, however, desire for more (Finke et al., 2019). The current approach to
reported meeting up with their friends in person much more addressing the desire of autistic people for more friends has
frequently than the autistic participants. It may be that autis- been to design intervention programs to help them learn the
tics find methods for interacting with their friends that do not requisite skills for making and maintaining friendships. The
require meeting in person, such as spending time with their skills targeted in these intervention programs are generally
friends playing video games. These data appear to continue based on neurotypical standards and expectations for friend-
to support the emerging idea that the opportunities provided ship behaviors and do not usually consider the emerging
by virtual interactions, and game-based interactions, could evidence on the discrepancies in the friendship preferences
provide an ideal context for friendship maintenance based on and practices between autistic and non-autistic people (Finke
the preferences reported by the autistic young adults in the et al., 2019).
current study (Finke, 2016; Finke et al., 2018, 2019). This Congruence, that is, similarities in perspectives and
is also consistent with the results for the survey item about beliefs particularly with respect to how to be a friend and
the proportion of friends the autistic participants had whom act in a friendship, is important to friendship intensity and
they had never met “in real life”. Autistic participants were longevity. Considering congruence in friendship perspec-
nearly five times more likely to report having friends they tives, beliefs, values, practices, and preferences may be the
had never met “in real life” (31%; n = 31/100) compared first step in designing intervention plans using a social model
to the non-autistic respondents (8.4%; n = 9/107). This sug- of disability lens. Helping an autistic person to choose an
gests virtual environments are important to establishing and appropriate peer to approach or to include in activities will
maintaining friendships for autistic people. need to focus on the “match” between autistics and their
potential friends. Higher levels of congruence on friend-
ship perspectives will likely increase the prospect that the

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3059

autistic person and the peer will get along and have positive Limitations
interactions with each other. Sustained and repeated positive
interactions are needed for friendship formation and main- Although this study utilized many best-practice guidelines
tenance (Finke, 2016). for survey development, some limitations affect the gener-
In addition to the findings reported from the current alizability of the results of the survey. First, respondents
project, autistic people have reported they would gener- were a self-selected convenience sample, not a randomized
ally prefer less physical and emotional closeness in their sample of the population; therefore, the respondents may
relationships. Autistics have also reported a preference for not be representative of all autistic and non-autistic young
being friends with people who share similar interests, that adults. It is possible that autistic people who had higher lev-
is, with people with whom they would have things to do. els of dissatisfaction with their current friendships, or who
This body of research offers a potential explanation for the felt making or maintaining friendships is very challenging
higher proportion of autistics who report being friends with for them, were more willing to commit the time to complete
other autistics (e.g., Bauminger & Shulman, 2003; Cook the survey than those who felt higher levels of satisfaction
et al., 2017). Autistic young adults have a higher likeli- with their current social outcomes.
hood of having congruent perspectives and preferences in Further, the demographic characteristics of the sample are
friendships with other autistics than with their non-autistic not be representative of the population as a whole. The pro-
peers. This information can be extremely helpful in guid- portion of autistic males compared to autistic females who
ing SLPs, teachers, parents, and others involved in the daily completed this survey was roughly consistent with overall
lives of autistic people in creating opportunities for interac- reported prevalence statistics. However, the proportion of
tions with others. These findings can also support clinical males and females across the groups was inverted. More
decision-making regarding using formats other than face-to- autistic males and more non-autistic females completed this
face interaction to scaffold and support friendship formation survey. This likely influenced the responses obtained from
and maintenance as these alternative modes of interaction the autistic and non-autistic groups and affected the patterns
may better accommodate the preferences of autistic people of results obtained. Finally, non-autistic participants were
(Finke, 2016). recruited largely using social media, therefore there was no
We are currently at the intersection of research a clinical way to confirm the participants in this group are truly with-
practice. In truth, we don’t yet know how the information out a diagnosis (that is that they were truly “non-autistic”).
learned in this research study, and others like it, translates All potential participants were asked whether they had an
to real life for autistic people or how these stated preferences ASD diagnosis or not, but diagnostic status was not con-
manifest during the friendship formation process. It is not firmed for the participants who were recruited through social
known if some of these preferences are non-negotiable or if media sources. Therefore, it is possible that a person with
all are negotiable to some extent in order to find common an undiagnosed autism spectrum disorder or a person with
ground with potential friend candidates. But what we do another diagnosis that could impact social interaction and
know is that autistic people have different friendship prefer- friendship behaviors could have participated in the “non-
ences and behave differently within their friendships than autistic” sample. This could affect the generalizability of the
many non-autistics. This needs to be the new basis for under- results to a truly “non-autistic” population.
standing the friendships of autistics and the new perspective Because the respondents were completing the survey on
with which researchers and clinicians alike think about, plan their own time and in spaces remote from the researchers,
for, and support autistic friendships. there were inherently limited opportunities for the partici-
None of these implications, or ways of thinking about pants to ask questions about the individual survey items.
autistic friendships as having different preferences and prac- For this reason, it is possible some participants interpreted
tices from neurotypical friendships, indicate autistic people the survey questions in different ways than other respond-
do not want or need to understand the norms or practices for ents, and that some of these interpretations varied from the
neurotypical friendships, however. Autistic people may want original intent of the researcher. A pilot of the survey instru-
to know what neurotypical people will expect in social situa- ment by autistic people was undertaken prior to distribution
tions. These implications simply mean that how friendships and revisions were made to the survey instrument based on
are constructed and maintained can occur along a broader feedback from the pilot participants prior to wide distribu-
spectrum than perhaps has been previously considered, and tion of the survey instrument to decrease the effect of this
that autistic preferred friendship behaviors are not wrong, limitation. Finally, it is important to acknowledge and report
just different. that the research team for this investigation was comprised
entirely of non-autistic adults.

13
3060 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Future Research Appendix 1

Future research should continue to examine the perspec- 1. Select the statement that mot applies to you:
tives, values, beliefs, and practices relative to friendships
of autistic people. Future research could collect data on a. When I talk with my friend(s) on the phone, it is
actual friendships or investigate specific friendship styles usually to make arrangements rather than to chat
or behaviors of autistic and non-autistic people. This infor- b. When I talk with my friend(s) on the phone, it is
mation would add to and expand on the current findings usually to chat rather than to make arrangements
on general friendship perspectives and preferences. Future
research could also investigate ways to support autistic 2. Select the statement that most applies to you:
people in identifying other individuals with congruent
friendship perspectives and preferences and examine how a. I prefer meeting my friend(s) for a specific activity,
to construct environments and contexts that may encourage e.g., going to the movies, playing golf
friendship formation for autistics. A better understanding of b. I prefer meeting my friend(s) for a chat, e.g., at a pub
the supports that may be necessary to scaffold emerging and or at a café
developing friendships for autistic young adults would help
address both the WHO model ‘disability’ and ‘impairment’ 3. Select the statement that most applies to you:
aspects of having and ASD diagnosis (Donaldson et al.,
2018). Finally, considering autistic friendship preferences a. My friend(s) value me more as someone who is a
in the way discussed in this paper will require an adjustment support to them than as someone to have fun with
in therapeutic processes and the ways assessments and inter- b. My friend(s) value me more as someone to have fun
ventions are conceived and executed. These adjustments and with than as someone who is a support to them
considerations should be investigated further through future
research investigations. 4. Select the statement that most applies to you:

a. If my friend(s) had a problem, I would be better

Conclusions at discussing their feelings about the problem than
coming up with practical solutions
In conclusion, the findings from the current project add b. If my friend(s) had a problem, I would be better at
additional information about the friendship preferences of coming up with practical solutions than discussing
autistic people and indicate autistic and non-autistic peo- their feelings about the problem
ple largely behave differently as friends. These patterns of
preferred friendship behavior are important to identifying 5. Select the statement that most applies to you:
and describing the culture of autistic friendship. Moving
forward, it seems the discussion about autistic friendships a. If my friend was having personal problems, I would
should be centered on autistic people as a distinct group wait for them to contact me as I wouldn’t want to
who do not and should not be expected to conform to neu- interfere
rotypical norms in constructing their social relationships. b. If my friend was having personal problems, I would
Differences between autistics and non-autistics have been contact them to discuss the problem
reported for as long as these comparisons have been made.
It is time the perspectives and experiences of autistic people 6. Select the statement that most applies to you:
are regarded as valid. Preferences and experiences that are
not the same as neurotypical standard norms for friendship a. If I have something critical to say to my friend(s), I
or not wrong and do not need to be changed. It is time to think it is best to broach the subject gently
reframe the interpretation of this information from seeing b. If I have something critical to say to my friend(s), I
differences as deficits to seeing differences as preferences think it is best to just come right out and say it
and using these preferences to support autistic people in hav-
ing the friendship outcomes and social lives they want for 7. If I fell out with my best friend and I thought that I
themselves. hadn’t done anything wrong, I would:

a. Do whatever it takes to repair the relationship

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3061

b. Be willing to make the first move, as long as they Variable B(SE) Wald (χ2) Sig OR
reciprocated
c. Be willing to sort out the problem, if they made the ASD 1.26(0.47) 7.32 0.007 3.54**
first move Gender 1.43(0.46) 9.71 0.002 4.16**
d. Not be able to be their close friend anymore ASD by Gender − 1.41(0.66) 4.63 0.031 0.24*

8. In terms of personality, how similar to your friend(s)

For Model: X(3) = 19.41, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001***
do you tend to be?
Model summary

a. Very dissimilar Step − 2 Log likelihood Cox & Snell R Nagelkerke R Square
b. Not very similar Square
c. Quite similar 1 260.30 0.09 0.12
d. Very similar
Hosmer and Lemeshow Test

9. In terms of interests, how similar to your friend(s) do Step Chi-square df Sig

you tend to be? 1 0.00 2 1.00
I would rather meet up with a friend to chat or for a specific activity
a. Very dissimilar
b. Not very similar
c. Quite similar Variable B(SE) Wald (χ2) Sig OR
d. Very similar ASD 1.56(0.47) 11.23 < 0.001 4.75***
Gender 0.99(0.45) 4.81 0.03 2.68*
10. How often do you make plans to meet with your ASD by Gender 0.04(0.68) 0.003 0.96 1.04
friend(s) that live near you?

a. Once or twice a year For Model: X(3) = 49.77, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001***
b. Once every 2–3 months
Model summary
c. Once a month
d. Once every couple of weeks Step − 2 Log likeli- Cox & Snell R Nagelkerke R
hood Square Square
e. Once or twice a week
f. 3 or 4 times a week 1 230.21 0.22 0.29
g. More than any of the above Hosmer and Lemeshow Test
Step Chi-square df Sig
11. How many friends to you have that you’ve never met
in “real life” or face-to-face? 1 0.00 2 1.00
My friends value me as a support or as someone to have fun with
a. Zero, I have been in close proximity during interac-
tions at some point with all of my friends Variable B(SE) Wald (χ2) Sig OR
b. One or two of my close friendships were made and
maintained without face-to-face interaction (e.g., ASD 1.31(0.49) 7.11 0.008 3.71**
only interacted online) Gender 1.39(0.49) 8.22 0.004 4.02**
c. Most of my close friendships were made and are ASD by Gender − 0.22(0.67) 0.10 0.75 0.81
maintained without face-to-face interaction (e.g.,
only interact online)
For Model: X(3) = 44.89, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001***
Model summary

Appendix 2 Step − 2 Log likeli- Cox & Snell R Nagelkerke R

hood Square Square
I would rather talk with a friend just to chat or to make 1 231.14 0.20 0.27
arrangements

13
3062 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Hosmer and Lemeshow Test Model summary

Step Chi-square df Sig Step − 2 Log likelihood Cox & Snell R Nagelkerke R Square
Square
1 0.000 2 1.00
1 243.62 0.14 0.18
I am better at discussing feelings or practical solutions
Hosmer and Lemeshow Test

Variable B(SE) Wald (χ2) Sig OR Step Chi-square df Sig

ASD 0.47(0.45) 1.12 0.29 1.61 1 0.000 2 1.00

Gender 1.16(0.45) 6.74 0.01 3.20** Binary Logistic Regression for Personality (similar or dissimilar)
ASD by Gender − 0.60(0.64) 0.88 0.35 0.55

Variable B(SE) Wald (χ2) Sig OR

For Model: X(3) = 11.90, p.008, p < 0.05*, p < 0.01**, p < 0.001***
Model summary ASD 1.01(0.57) 3.11 0.08 2.74
Gender 0.77(0.59) 1.70 0.19 2.16
Step − 2 Log likeli- Cox & Snell R Nagelkerke R
hood Square Square ASD by gender − 0.36(0.78) 0.22 0.64 0.70

1 262.51 0.06 0.08 For Model: X(3) = 11.30, p.010, p < 0.05*, p < 0.01**, p < 0.001***
Hosmer and Lemeshow Test Model summary

Step Chi-square df Sig Step − 2 Log likelihood Cox & Snell R Nagelkerke R Square
Square
1 0.000 2 1.00
1 202.88 0.05 0.08
I would contact my friend to discuss a problem or I would wait for
them to contact me Hosmer and Lemeshow Test
Step Chi-square df Sig
2
Variable B(SE) Wald (χ ) Sig OR 1 0.000 2 1.00
ASD 1.39(0.48) 8.52 0.004 4.03** Binary Logistic Regression for Interests (similar or dissimilar)
Gender − 0.37(0.47) 0.64 0.43 0.69
ASD by Gender − 70(0.69) 1.02 0.31 2.01
Variable B(SE) Wald (χ2) Sig OR

ASD 0.07(0.87) 0.006 0.94 1.07

Note. For Model: X(3) = 33.43, p. < 001, p < 0.05*, p < 0.01**, p < 0.001*** Gender 0.47(0.76) 0.38 0.54 1.60
Model summary ASD by Gender 1.25(1.09) 1.31 0.25 3.47
Step − 2 Log likelihood Cox & Snell R Nagelkerke R Square
Square
For Model: X(3) = 17.16, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001***
1 239.78 0.16 0.21 Model summary
Hosmer and Lemeshow test Step − 2 Log likeli- Cox & Snell R Nagelkerke R
Step Chi-square df Sig hood Square Square

1 0.000 2 1.00 1 157.99 0.08 0.14

If I had something to say to a friend I would just come out and say it Hosmer and Lemeshow Test
or I would broach the subject gently Step Chi-square df Sig

1 0.000 2 1.00
Variable B(SE) Wald (χ2) Sig OR
How often do you meet up with your friends who live near you
ASD − 0.03(0.51) 0.003 0.95 0.97 (“weekly” or “monthly or less”)
Gender − 0.77(0.46) 2.78 0.10 2.16
ASD by Gender 1.03(0.68) 2.26 0.13 2.79

For Model: X(3) = 29.22, p. < 001, p < 0.05*, p < 0.01**, p < 0.001***

13
Journal of Autism and Developmental Disorders (2023) 53:3047–3064 3063

Variable B(SE) Wald (χ2) Sig OR References

ASD 2.19(0.52) 18.02 < 0.001 8.96*** Bagwell, C. L., & Schmidt, M. E. (2011). Friendships in childhood and
Gender − 0.44(0.70) 0.40 0.53 0.64 adolescence. Guilford Press.
ASD by Gender 0.04(0.84) 0.002 0.97 0.1.04 Baron-Cohen, S., & Wheelwright, S. (2004). The empathy quotient: An
investigation of adults with Asperger syndrome or high function-
ing autism, and normal sex differences. Journal of Autism and
For Model: X(3) = 38.31, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001*** Developmental Disorders, 34(2), 163–175.
Model Summary Bauminger, N., & Kasari, C. (2000). Loneliness and friendship in
high-functioning children with autism. Child Development, 71,
Step -2 Log likeli- Cox & Snell R Nagelkerke R 447–456.
hood Square Square Bauminger, N., & Shulman, C. (2003). The development and main-
tenance of friendship in high-functioning children with autism:
1 203.91 0.18 0.25 Maternal perceptions. Autism, 7, 81–97.
Hosmer and Lemeshow Test Bauminger, N., Solomon, M., Aviezer, A., Heung, K., Brown, J., &
Rogers, S. J. (2008a). Friendship in high-functioning children with
Step Chi-square df Sig autism spectrum disorder: Mixed and non-mixed dyads. Journal
of Autism and Developmental Disorders, 38(7), 1211–1229.
1 0.000 2 1.00 Bauminger, N., Solomon, M., Aviezer, A., Heung, K., Gazit, L., Brown,
How many close friends do you have that you’ve never met “in real J., & Rogers, S. J. (2008b). Children with autism and their friends:
life” or face-to-face (zero or some A multidimensional study of friendship in high-functioning
autism spectrum disorder. Journal of Abnormal Child Psychol-
ogy, 36, 135–150.
Variable B(SE) Wald (χ2) Sig OR Blume, H. (1998). Neurodiversity. The Atlantic. Retrieved January 5,
2022 from http://w ​ ww.t​ heatl​ antic.c​ om/m
​ agazi​ ne/a​ rchiv​ e/1​ 998/0​ 9/​
ASD 1.66(0.56) 8.90 0.003 5.26** neuro​diver​sity/​305909/.
Gender − 0.34(0.83) 0.16 0.69 0.71 Bolis, D., Lahnakoski, J. M., Seidel, D., Tamm, J., & Schilbach, L.
(2021). Interpersonal similarity of autistic traits predicts friend-
ASD by Gender 0.11(0.96) 0.01 0.91 1.12 ship quality. Social Cognitive and Affective Neuroscience, 16(1–
2), 222–231.
Brownlow, C., Bertilsdotter Rosqvist, H., & O’Dell, L. (2015). Explor-
For Model: X(3) = 18.02, p < 0.001, p < 0.05*, p < 0.01**, p < 0.001*** ing the potential for social networking among people with autism:
Challenging dominant ideas of ‘friendship.’ Scandinavian Journal
Model Summary of Disability Research, 17(2), 188–193.
Calder, L., Hill, V., & Pellicano, E. (2013). ‘Sometimes I want to play
Step − 2 Log likeli- Cox & Snell R Nagelkerke R by myself’: Understanding what friendship means to children with
hood Square Square autism in mainstream primary schools. Autism, 17(3), 296–316.
Carrington, S., Templeton, E., & Papinczak, T. (2003). Adolescents
1 185.20 0.08 0.13 with Asperger syndrome and perceptions of friendship. Focus on
Autism and Other Developmental Disabilities, 18(4), 211–218.
Hosmer and Lemeshow Test
Carter, E. W. (2016). A place of belonging: Research at the intersection
Step Chi-square df Sig of faith and disability. Review & Expositor, 113, 167–180.
Carter, E. W. (2021). Dimensions of belonging for individuals with
1 0.000 2 1.00 intellectual and developmental disabilities. In J. L. Jones, & K.
L. Gallus (Eds.), Belonging and Resilience in individuals with
developmental disabilities (pp. 39–78). Springer, Cham.
Acknowledgements The authors would like to thank the Interactive
Chen, Y. L., Senande, L. L., Thorsen, M., & Patten, K. (2021). Peer
Autism Network for their assistance in recruiting the sample of young
preferences and characteristics of same-group and cross-group
adults with ASD. The authors would also like to thank all of the young
social interactions among autistic and non-autistic adolescents.
adults who participated in this research study.
Autism, 25(7), 1885–1900.
Clark, M. L., & Ayers, M. (1993). Friendship expectations and friend-
Author Contributions Material preparation, data collection and analy- ship evaluations: Reciprocity and gender effects. Youth & Society,
sis were performed by EHF. All drafts and revisions of the manuscript 24(3), 299–313.
were written by EHF. Cook, A., Ogden, J., & Winstone, N. (2017). Friendship motivations,
challenges and the role of masking for girls with autism in con-
Declarations trasting school settings. European Journal of Special Needs Edu-
cation, 33, 1–14.
Creswell, J. W. (1998). Qualitative inquiry and research design:
Conflict of interest The authors do not have any actual or potential
Choosing among five traditions. Sage.
conflict of interest to report. This includes financial, personal or other
Crompton, C. J., Hallett, S., Ropar, D., Flynn, E., & Fletcher-Watson,
relationships with other people or organizations that could inappropri-
S. (2020). ‘I never realised everybody felt as happy as I do when I
ately influence, or be perceived to influence, our work.
am around autistic people’: A thematic analysis of autistic adults’
relationships with autistic and neurotypical friends and family.
Autism, 24(6), 1438–1448.

13
3064 Journal of Autism and Developmental Disorders (2023) 53:3047–3064

Daniel, L. S., & Billingsley, B. S. (2010). What boys with an autism Maslow, A. (1954). Motivation and personality. Harpers.
spectrum disorder say about establishing and maintaining friend- Mendelson, J. L., Gates, J. A., & Lerner, M. D. (2016). Friendship in
ships. Focus on Autism and Other Developmental Disabilities, school-age boys with autism spectrum disorders: A meta-analytic
25(4), 220–229. summary and developmental, process-based model. Psychological
Dean, M., Harwood, R., & Kasari, C. (2017). The art of camouflage: Bulletin, 142(6), 601–622.
Gender differences in the social behaviors of girls and boys with Milton, D. E. (2012). On the ontological status of autism: The ‘double
autism spectrum disorder. Autism, 21(6), 678–689. empathy problem.’ Disability & Society, 27(6), 883–887.
Dean, M., Kasari, C., Shih, W., Frankel, F., Whitney, R., Landa, R., Milton, D., Heasman, B., & Sheppard, E. (2018). Double empathy.
& Harwood, R. (2014). The peer relationships of girls with ASD In F. Volkmar (Ed.), Encyclopedia of autism spectrum disorders
at school: Comparison to boys and girls with and without ASD. (pp. 1–8) Springer.
Journal of Child Psychology and Psychiatry, 55(11), 1218–1225. Morrison, K. E., DeBrabander, K. M., Jones, D. R., Faso, D. J., Ack-
DeRosier, M. E., Swick, D. C., Davis, N. O., McMillen, J. S., & Mat- erman, R. A., & Sasson, N. J. (2020). Outcomes of real-world
thews, R. (2011). The efficacy of a social skills group intervention social interaction for autistic adults paired with autistic compared
for improving social behaviors in children with high functioning to typically developing partners. Autism, 24(5), 1067–1080.
autism spectrum disorders. Journal of Autism and Developmental Nicolaidis, C., Raymaker, D., McDonald, K., Dern, S., Boisclair, W.
Disorders, 41(8), 1033–1043. C., Ashkenazy, E., & Baggs, A. (2013). Comparison of healthcare
Donaldson, A. L., Nolfo, M., & Montejano, M. (2018). Relationships, experiences in autistic and non-autistic adults: A cross-sectional
friendships, and successful social communication: Addressing online survey facilitated by an academic-community partnership.
disability. Seminars in Speech and Language, 39, 166–177. Journal of General Internal Medicine, 28(6), 761–769.
Felmlee, D., Sweet, E., & Sinclair, H. C. (2012). Gender rules: Same- Parker, J. G., & Asher, S. R. (1993). Friendship and friendship qual-
and cross-gender friendships norms. Sex Roles, 66(7), 518–529. ity in middle childhood: Links with peer group acceptance and
Finke, E. H. (2016). Friendship: Operationalizing the intangible to feelings of loneliness and social dissatisfaction. Developmental
improve friendship-based outcomes for individuals with autism Psychology, 29(4), 611.
spectrum disorder. American Journal of Speech-Language Pathol- Parsons, S. (2015). ‘Why are we an ignored group?’ Mainstream edu-
ogy, 25(4), 654–663. cational experiences and current life satisfaction of adults on the
Finke, E. H., Hickerson, B. D., & Kremkow, J. M. (2018). “To be quite autism spectrum from an online survey. International Journal of
honest, if it wasn’t for videogames I wouldn’t have a social life at Inclusive Education, 19(4), 397–421.
all”: Motivations of young adults with autism spectrum disorder Petrina, N., Carter, M., & Stephenson, J. (2014). The nature of friend-
for playing videogames as leisure. American Journal of Speech- ship in children with autism spectrum disorders: A systematic
Language Pathology, 27(2), 672–689. review. Research in Autism Spectrum Disorders, 8, 111–126.
Finke, E. H., McCarthy, J. H., & Sarver, N. A. (2019). Self-perception Petrina, N., Carter, M., Stephenson, J., & Sweller, N. (2016). Perceived
of friendship style: Young adults with and without autism spec- friendship quality of children with autism spectrum disorder as
trum disorder. Autism & Developmental Language Impairments, compared to their peers in mixed and non-mixed dyads. Journal of
4, 2396941519855390. Autism and Developmental Disorders, 46(4), 1334–1343.
Hall, J. A. (2012). Friendship standards: The dimensions of ideal Rea, L. M., & Parker, R. A. (2005). Designing and constructing survey
expectations. Journal of Social and Personal Relationships, 29(7), research. San Francisco.
884–907. Sedgewick, F., Hull, L., & Ellis, H. (2021). Autism and masking: How
Hall, J. A., Larson, K. A., & Watts, A. (2011). Satisfying friendship and why people do it, and the impact it can have. Jessica Kingsley
maintenance expectations: The role of friendship standards and Publishers.
biological sex. Human Communication Research, 37(4), 529–552. Singer, J. (2017). Neurodiversity: The birth of an idea.
Hyman, L., Lamb, J., & Bulmer, M. (2006, April). The use of pre-exist- Smith, A., & Anderson, M. (2018). Social media use in 2018.
ing survey questions: Implications for data quality. In Proceedings Solomon, M., Goodlin-Jones, B. L., & Anders, T. F. (2004). A social
of the European Conference on Quality in Survey Statistics (pp. adjustment enhancement intervention for high functioning autism,
1–8). Wales, UK: Cardiff. Asperger’s syndrome, and pervasive developmental disorder NOS.
Kapp, S. K., Gillespie-Lynch, K., Sherman, L. E., & Hutman, T. Journal of Autism and Developmental Disorders, 34(6), 649–668.
(2012). Deficit, difference, or both? Autism and neurodiversity. van Selm, M., & Jankowski, N. W. (2006). Conducting online surveys.
Developmental Psychology, 49, 59–71. Quality and Quantity, 40(3), 435–456.
Kuo, M. H., Orsmond, G. I., Cohn, E. S., & Coster, W. J. (2013). Vine Foggo, R. S., & Webster, A. A. (2017). Understanding the social
Friendship characteristics and activity patterns of adolescents with experiences of adolescent females on the autism spectrum.
an autism spectrum disorder. Autism, 17(4), 481–500. Research in Autism Spectrum Disorders, 35, 74–85.
Kuo, M. H., Orsmond, G. I., Coster, W. J., & Cohn, E. S. (2014). Webster, A. A., & Carter, M. (2010). Adaptation of an interview-
Media use among adolescents with autism spectrum disorder. based protocol to examine close relationships between children
Autism: The International Journal of Research and Practice, with developmental disabilities and peers. Australasian Journal
18(8), 914–923. of Special Education, 34, 79–94.
Laugeson, E. A., Frankel, F., Gantman, A., Dillon, A. R., & Mogil, C. Wiseman, J. P. (1986). Friendship: Bonds and binds in a voluntary
(2012). Evidence-based social skills training for adolescents with relationship. Journal of Social and Personal Relationships, 3(2),
autism spectrum disorders: The UCLA PEERS program. Journal 191–211.
of Autism and Developmental Disorders, 42(6), 1025–1036.
Locke, J., Ishijima, E. H., Kasari, C., & London, N. (2010). Loneliness, Publisher's Note Springer Nature remains neutral with regard to
friendship quality and the social networks of adolescents with jurisdictional claims in published maps and institutional affiliations.
high-functioning autism in an inclusive school setting. Journal of
Research in Special Educational Needs, 10, 74–81.
Long, J., Brown, J., Daly, S., Gibson, K., & McNeillis, C. (2018).
Friendship and sociality in autism services. Good Autism Practice
(GAP), 19(1), 22–31.

13
Journal of Autism & Developmental Disorders is a copyright of Springer, 2023. All Rights
Reserved.