The Association Between Self-Reported Camouflaging of Autistic Traits and Social Competence in Nonautistic Young Adults

Abstract

Background:

Autistic individuals often experience difficulties in social settings. Although autistic individuals may not intuitively know the “typical” way to behave in social settings, many autistic individuals have a desire to fit in so they develop techniques to “camouflage” their autistic traits. Although camouflaging may help individuals to navigate social environments, camouflaging has also been shown to produce negative psychological outcomes. This study aims to explore whether this “camouflaging” strategy is associated with poor social competence, an aspect of the autism diagnosis.

Methods:

In this study, 247 nonautistic adults completed the Multidimensional Social Competence Scale (MSCS) to assess their social competence, and the Camouflaging Autistic Traits Questionnaire (CAT-Q) to assess the extent to which they used strategies to compensate or mask behaviors characteristic of autism in social settings.

Results:

We found that over and above IQ, gender, and executive functioning scores, social competence (MSCS) scores reliably predicted the extent to which nonautistic individuals camouflaged, accounting for 25% of the variance in CAT-Q scores. Importantly, even when autistic traits were controlled for, social competence was still able to account for additional variance in CAT-Q scores.

Conclusion:

These results suggest that low social competency in nonautistic adults predicts camouflaging as a strategy in social situations. Given these camouflaging behaviors are being performed in an attempt to comply with an environmental demand to behave in a particular manner, these results also highlight the importance of conceptualizing the social challenges that autistic and nonautistic individuals face in a bidirectional manner, where the onus is not solely on the individual to comply with social conventions but also on society to accommodate diverse behavioral traits.

Lay summary

Why was this study done?

Some autistic individuals try to hide their autistic traits to “fit in” with others, referred to as “camouflaging.” Nonautistic adults also report camouflaging, but it is unclear whether this camouflaging is related to social difficulties that are not specific to autism. No research has been conducted to examine the relationship between social competence and camouflaging in nonautistic adults.

What was the purpose of this study?

To further understand the factors that are related to camouflaging behaviors. More specifically, whether social abilities, and/or autism characteristics, are related to whether nonautistic adults camouflage.

What did the researchers do?

We had 257 nonautistic adults complete various questionnaires, including ones that asked them about their camouflaging behaviors and social abilities. We examined the relationships between the scores from these questionnaires and the influence of other factors such as gender, intelligence, and executive functioning.

What were the results of the study?

We found that both social abilities and autistic traits were related to camouflaging behaviors. Indeed, nonautistic adults who had poor social skills, and more autistic traits, engaged in more camouflaging. Social skills were associated with camouflaging even after we considered factors such as gender, intelligence, and executive functioning.

What do these findings add to what was already known?

These findings help us understand camouflaging by demonstrating that it may be a common response to social difficulties in nonautistic, as well as autistic, adults. These results also indicate that camouflaging is related to low social competency, not just autism characteristics.

What are the potential weaknesses in the study?

The participants in our study completed questionnaires through which they were required to pick from set answers, rather than describe their experiences. We may be missing important qualitative differences in the way nonautistic adults camouflage compared with autistic adults.

How will these findings help autistic adults now or in the future?

By comparing what is shared and what is unique with nonautistic people who share traits with autistic people, a more precise definition and study of camouflaging behavior are possible. Rather than see camouflaging as a phenomenon that occurs exclusively in autistic people because of their disability, it may be that both autistic and nonautistic people use camouflaging when they perceive themselves to lack the necessary social competencies that are expected within their social contexts. Because both autistic traits and social competency are related to camouflaging behavior, we can begin to think about how to tease apart which characteristics are more likely to evoke camouflaging in autistic individuals and how this may be similar or different in nonautistic individuals. This knowledge will ultimately contribute to the development of more tailored approaches to prevent and/or reduce the negative impact of camouflaging behaviors for autistic adults.

Introduction

When interacting with others in social settings, being yourself is not always the most successful approach. This may be particularly true for autistic individuals. Autism spectrum disorder (ASD) is a developmental disorder characterized by persistent difficulties in social communication and interaction, as well as restricted and repetitive patterns of behavior.¹ The social communication and interaction difficulties experienced by autistic individuals often relate to differences in social behavior, which may not match with what would typically be considered the most socially appropriate way to interact, despite social expectations being relatively ambiguous and frequently changing across diverse social settings. Many autistic individuals have less interest in or ability to engage in social chit-chat, leading others to assume they are uninterested in socializing.^2,3 Autistic individuals often display repetitive behaviors such as “stimming,” or self-stimulatory movements such as hand or finger flapping aimed at shutting out distressing external stimuli.⁴ These behaviors can cause observers to develop a negative perception of the autistic individual, even after very little exposure.⁵ Although autistic individuals may not intuitively know the “typical” way to behave in many social settings, many autistic individuals have the desire to fit in and develop techniques to “camouflage” their autistic traits in an attempt to appear nonautistic.^6–10

Social camouflaging is defined as the initiation of particular behaviors in social settings in an attempt to hide less socially normative traits or behaviors, such as traits characteristic of autism.^6–10 Autistic individuals frequently camouflage in an attempt to “mask” or “compensate” for their autistic traits.^6,10–12 To do so, autistic individuals will often observe their peers and mimic or copy their behaviors.^6,13,14 For example, autistic individuals may force themselves to make eye contact and show displays of emotion while interacting with others, despite feeling uncomfortable making eye contact and feeling like their emotions are inauthentic.^6,10 Autistic individuals may also suppress their “stims” in an attempt to avoid drawing negative attention toward themselves, despite the fact that “stimming” helps autistic individuals to find relief from excessive sensory stimulation and feelings of anxiety.^4,15 Although camouflaging may be a coping strategy that helps autistic individuals navigate difficult social environments,¹² many negative side effects of camouflaging have been reported.^6–8,16,17

Many autistic individuals see camouflaging as a way to make friends and form romantic attachments, however, there are harmful effects of camouflaging.⁶ Autistic individuals report that when they camouflage, it influences their self-identity, making it difficult for them to keep track of their authentic self.⁶ Autistic individuals also report that camouflaging is exhausting, as it feels like they are constantly putting on an act.^6,14 Given that these camouflaging behaviors are effortful and are not always successful, frequent camouflaging is also associated with high levels of stress and low self-esteem.^6,8 Ineffective camouflaging is associated with poor outcomes; however, effective camouflaging may be just as harmful. Since effective camouflaging may allow autistic individuals to fit in, it can lead to the misperception that an autistic individual is functioning well. Since autistic individuals do not appear to be struggling, they may not receive the support and resources they need to effectively manage their autistic traits, likely resulting in high rates of stress, anxiety, and depression.^6,7,16,17

It is clear that many autistic individuals camouflage in an attempt to improve their success in social situations. It has been reported that camouflaging is extremely effortful and may put autistic individuals at risk for psychological harm.^6,18 It has also been reported that when nonautistic individuals behave in a manner that is inconsistent with their innate tendencies to accommodate society's expectations, such as introverts who try to act more extraverted, they are at risk for psychological harm.^19,20 Thus, it appears that camouflaging is something that autistic and nonautistic people do to try to manage the way other people perceive them, or to hide the less socially acceptable aspects of their social behavior. It has been reported that nonautistic adults also camouflage, and the extent to which these nonautistic adults camouflage is related to their autistic traits and social anxiety.⁸ However, given that social difficulties are an inherent part of the autism diagnosis, it is unclear if social competence on its own is related to camouflaging behaviors.

It is important to understand if camouflaging strategies used to cope with social challenges are associated with social competence. Compared with autistic individuals, camouflaging strategies may be more commonly adopted by nonautistic individuals who have lower social competency, but not necessarily high autistic traits. The aim of this study is to investigate whether social competence, measured by the Multidimensional Social Competence Scale (MSCS),²¹ can predict camouflaging behaviors in a nonautistic sample, as measured by the Camouflaging Autistic Traits Questionnaire (CAT-Q).⁸ Participants also completed the Behavior Rating Inventory of Executive Function for Adults (BRIEF-A) to measure their executive functioning, as previous research has shown that executive functioning is related to social abilities.^22,23 If camouflaging is a behavior adopted by all individuals with poor social competence, we expect to find a strong association between camouflaging behaviors and the social competence of nonautistic individuals.

Methods

Participants

Two hundred forty-seven participants between the ages of 17 and 34 years (M = 20.03, SD = 2.28; 176 female) participated in this study. Participants provided informed consent and received course credit for their participation. All procedures were approved by the Simon Fraser University Research Ethics Board, and were in accordance with the World Medical Association 2013 Declaration of Helsinki.

Procedures

Participants were recruited from introductory university psychology classes. All participants self-reported as nonautistic. Following informed consent, participants completed the Wechsler Abbreviated Scales of Intelligence II (WASI-II) with a research assistant,²⁴ then a short demographic questionnaire followed by the four self-report questionnaires (see the Measures section). The study took ∼60 minutes to complete.

Measures

Autism Spectrum Quotient

The Autism Spectrum Quotient is a 50-item self-report measure developed to quantify autism characteristics.²⁵ The AQ covers five domains associated with autism, including social skills, imagination, attention to detail, communication skills, and attention switching/tolerance to change. Individuals who score 32+ are considered to have high autistic traits, and generally speaking, higher scores are indicative of poor social skills, imagination, communication skills, and attention switching/tolerance to change, and exceptional attention to detail.

Multidimensional Social Competence Scale

To assess social competence, participants completed the self-report MSCS.²¹ The MSCS was created to measure the heterogeneity in social competence in autistic and nonautistic individuals. The MSCS has been validated as a self-report measure of social competence for individuals older than 17 years.²⁶ The MSCS consists of 77 questions utilizing a 5-point Likert scale, with higher total scores of the MSCS indicating better social competence.

The MSCS differentiates various components of social competence into seven distinct domains: social motivation, social inferencing, demonstrating empathic concern, social knowledge, verbal conversation skills (VCS), nonverbal sending skills, and emotional regulation.

The self-report MSCS utilized in this study was validated with 1178 nonautistic young adults (360 males) ranging in age from 17 to 25 years (M = 19.72, SD = 1.61). Trevisan et al. identified that the internal consistency for the full scale of the MSCS (α = 0.795) is good, and the seven domains and full scale of the MSCS range from acceptable (α = 0.739) to good (α = 0.826).²⁶ While the MSCS is not a measure of ASD symptoms, the MSCS total score shows good convergent validity with the AQ total score (r = 0.766), likely due to the AQ subscales of social skills and communication skills that specifically relate to dimensions of social competence.²⁶ Specifically, the social skills subscale shows acceptable convergent validity with the VCS domain of the MSCS (r = 0.68), while the communication skills subscale of the AQ shows acceptable convergent validity with the VCS domain of the MSCS (r = 0.59).

Camouflaging Autistic Traits Questionnaire

The CAT-Q is a self-report questionnaire measuring behaviors and strategies implemented to camouflage autistic traits using a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree).⁸ Respondent's answers across 25 questions yield three factor scores: Compensation (compensatory strategies to offset perceived social and communication difficulties), Masking (strategies utilized to pass as or appear nonautistic to other people), and Assimilation (strategies for blending in with social situations where the respondent is uncomfortable). The three factor scores are summed to create the total CAT-Q score, where higher scores represent a greater level of camouflaging. The CAT-Q was validated with samples of autistic and nonautistic adults older than 16 years, where total CAT-Q scores are positively correlated with autistic traits in both samples. The CAT-Q has high internal consistency (α = 0.94) and acceptable test/retest reliability (r = 0.77).

Behavior Rating Inventory of Executive Function–Adult Version

The BRIEF-A is a self-report questionnaire that measures executive functioning in adults between the ages of 18 and 90 years.²² The BRIEF-A consists of 75 questions in which respondents indicate how often they have exhibited the listed behaviors or problems within the last month using a 3-point scale (1 = never, 2 = sometimes, 3 = often). The BRIEF-A measures distinct aspects of executive functioning using nine clinical scales: Inhibit, Shift, Emotional Control, Self-Monitoring, Initiate, Working Memory, Plan/Organize, Task Monitor, and Organization of Materials. The Metacognitive Index (MI) is composed of the Initiate, Working Memory, Plan/Organize, Task Monitor, and Organization of Materials subscales, and represents the respondent's ability to cognitively manage attention for planning, organization, and problem solving. Given the established relationships between executive functioning and social skills,^22,23 for the purposes of this study we use the MI, which provides an index of executive functioning. The MI is found to have a high internal consistency (α = 0.94) with a normative sample of clinically healthy adults.²² Similarly, the MI has high test/retest reliability (r = 0.93).

Wechsler Abbreviated Scales of Intelligence II

The cognitive ability of all participants was assessed using the WASI-II.²⁴ The WASI-II provides an accurate estimate of cognitive ability in individuals, ages from 6 to 90 years, comparable with other measures of intelligence. The Full-Scale IQ-2 (FSIQ-2) was administered and scored by trained research assistants in the current study. The FSIQ-2 scale on the WASI-II is composed of scores from two subtests: Vocabulary and Matrix Reasoning. The Vocabulary subtest prompts examinees to define words that are presented visually and orally by the examiner, and assesses the participant's word knowledge, verbal concept formation, and crystallized intelligence. The Matrix Reasoning subtest has examinees select a correct picture response to complete matrices presented to them from the WASI-II stimulus book. This subtest assesses fluid intelligence, spatial ability, and perceptual organization, among other nonverbal reasoning abilities.

Results

Descriptive analyses

Table 1 presents descriptive statistics for each of the variables in this study.

Table 1.

Descriptive Statistics

Variable	Mean (SD; range)
Age	20.03 (2.28; 17.25–34.41)
IQ	100.62 (8.459; 85–129)
AQ	17.02 (6.35; 3–34)
MSCS	297.46 (26.31; 226–360)
CAT-Q	91.21 (23.34; 39–163)
BRIEF-MI	54.60 (9.86; 36–85)

AQ, Autism Quotient; BRIEF-MI, Behavior Rating Inventory of Executive Functioning–Metacognitive Index; CAT-Q, Camouflaging Autistic Traits Questionnaire—total score; IQ, intelligence quotient; MSCS, Multidimensional Social Competence Scale—total score.

Correlational analyses

Table 2 presents a correlational matrix depicting the relationships between each of the variables in the study. With the exception of a marginal association between the imagination subscale of the AQ and the VCS subscale of the MSCS, the social skills, attention switching/tolerance to change, communication skills, and imagination subscales of the AQ were significantly negatively correlated (all p ≤ 0.006) with each of the subscales of the MSCS. These negative correlations indicate that difficulties in these domains are associated with poorer ratings of social competence. Notably, the attention to details subscale of the AQ was not correlated with any of the subscales of the MSCS.

Table 2.

Correlations Among All Variables

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Sex	—
2. IQ	−0.042	—
3. BRIEF-MI	−0.065	0.050	—
4. AQ social skills	−0.058	−0.027	0.263**	—
5. AQ attention switching	−0.081	0.007	0.348**	0.463**	—
6. AQ attention to details	0.005	−0.155	−0.113	−0.014	0.039	—
7. AQ communication skills	−0.078	−0.009	0.307**	0.631**	0.468**	−0.001	—
8. AQ imagination	−0.158*	−0.153*	0.082	0.308**	0.201**	−0.024	0.328**	—
9. MSCS social motivation	0.104	−0.031	−0.323**	−0.732**	−0.429**	0.048	−0.635**	−0.247**	—
10. MSCS social inferencing	0.042	0.151*	−0.349**	−0.409	−0.281**	0.052	−0.460**	−0.231**	0.356**	—
11. MSCS demonstrating empathic concern	0.247**	0.048	−0.145*	−0.413**	−0.214**	0.075	−0.399**	−0.274**	0.490**	0.336**	—
12. MSCS social knowledge	0.237**	0.059	−0.189**	−0.240**	−0.179*	0.080	−0.339**	−0.173*	0.241**	0.472**	0.449**	—
13. MSCS verbal communication skills	0.147*	0.074	−0.358**	−0.194**	−0.436**	−0.052	−0.447**	−0.111	0.202**	0.342**	0.151*	0.263**	—
14. MSCS nonverbal communication skills	0.240**	0.130*	−0.321**	−0.488**	−0.322**	−0.059	−0.462**	−0.326**	0.580**	0.430**	0.621**	0.467**	0.242**	—
15. MSCS emotional regulation	−0.044	0.169*	−0.329**	−0.305**	−0.472**	−0.082	−0.408**	−0.183*	0.258**	0.352**	0.048	0.201*	0.588**	0.266**	—
16. CATQ total score	−0.041	0.333**	−0.038	0.443**	0.413**	0.065	0.509**	0.184*	−0.503**	−0.380**	−0.251**	−0.112	−0.268**	−0.414**	−0.337**	—

n = 247.

^* p ≤ 0.05.

^** p ≤ 0.001.

Regression analyses

Regression analyses were conducted to investigate the relationship between social competency (MSCS) and camouflaging behaviors (CAT-Q). Given it has been reported that camouflaging in nonautistic adults is related to their autistic traits,⁸ we also investigated the relation between autistic traits (AQ) and camouflaging behaviors. Three hierarchical multiple linear regression analyses were conducted to determine whether (1) self-reported social competence (MSCS) scores and (2) autistic traits (AQ) could account for additional variance in camouflaging (CAT-Q) scores, over and above IQ, sex, and executive functioning (BRIEF-MI), and (3) whether social competence scores (MSCS) could account for additional variance in camouflaging behaviors (CAT-Q) over and above autistic traits (AQ). See Tables 3–5, respectively, for full details on each regression model.

Table 3.

Hierarchical Regression Analysis Predicting Camouflaging Using Social Competence

Variable	CAT-Q
	Model 1		Model 2
	B	β	B	β
Constant	64.327		187.847
IQ	−0.154	−0.056	0.036	0.013
Sex	−1.132	−0.022	−0.014	0.001
BRIEF-MI	0.790**	0.334	0.203	0.086
MSCS social motivation			−1.094**	−0.317
MSCS social inferencing			−0.863*	−0.199
MSCS demonstrating empathic concern			0.092	0.021
MSCS social knowledge			0.910*	0.182
MSCS verbal conversation skills			−0.168	−0.041
MSCS nonverbal sending skills			−0.710*	−0.172
MSCS emotional regulation			−0.490	−0.127
R 2	0.114		0.363
F	10.429**		13.439**
ΔR2	—		0.249
ΔF	—		13.164**

n = 247.

^* p < 0.05.

^** p < 0.001.

Table 5.

Hierarchical Regression Analysis Predicting Camouflaging Using Autistic Traits and Social Competence

Variable	CAT-Q
	Model 1		Model 2		Model 3
	B	β	B	β	B	β
Constant	64.327		35.336		105.842
IQ	−0.154	−0.056	−0.001	>0.001	0.054	0.020
Sex	−1.132	−0.022	1.126	0.022	2.599	0.050
BRIEF-MI	0.790**	0.334	0.480**	0.203	0.332*	0.140
AQ (total)	—	—	1.672**	0.462	1.191**	0.330
MSCS (total)	—	—	—	—	−0.204*	−0.230
R 2	0.114		0.306		0.330
F	10.429**		26.672**		23.736**
ΔR2	—		0.192		0.024
ΔF	—		15.957**		8.628

n = 247.

^* p < 0.05.

^** p < 0.001.

Additional tests were conducted to ensure that the models did not violate the assumptions of the hierarchical multiple linear regression test. The residuals were determined to be independent, as indicated by a Durbin–Watson statistic of 2.27, 2.30, and 2.34, respectively. We also determined our data were homoscedastic by visually inspecting a plot of standardized residuals versus unstandardized predicted values. Tolerance values were calculated for each of the predictor variables in the models. The lowest values were 0.431, 0.535, and 0.454, respectively, and thus, there was no evidence of multicollinearity. Standard residual values were all between ±3 and the maximum Cook's distances were 0.08, 0.04, and 0.05, suggesting there were no influential observations that may skew the calculated values. Finally, normality of the residuals was verified by examining the Q-Q plots.

In the first step of the model (Model 1), IQ, sex, and executive functioning were determined to significantly predict CAT-Q scores, F(3, 243) = 10.429, p < 0.001. The addition of self-reported social competence scores to the prediction of CAT-Q scores (Model 2) led to a statistically significant increase in R² of 0.249, F(7, 236) = 13.164, p < 0.001, with lower social competence scores associated with increased camouflaging. Specifically, the social motivation, social inferencing, social knowledge, and nonverbal sending skills subscales all contributed significantly to the model (p < 0.03), while the emotional regulation subscale approached significance (p = 0.07; Table 3). The full model that used IQ, sex, executive functioning, and the subscales of the MSCS were able to predict CAT-Q scores (Model 2) with statistical significance, F(10, 236) = 13.439, p < 0.001; R² = 0.363.

To examine the association between autism characteristics and CAT-Q scores, we started with Model 1 (see above). Self-report of autism characteristics scores were then added to the prediction of CAT-Q scores (Model 2), leading to a statistically significant increase in R² of 0.222, F(5, 238) = 15.957, p < 0.001, with higher AQ scores associated with increased camouflaging. Specifically, the social skills, attention switching/tolerance to change, and communication skills subscales all contributed significantly to the model (p < 0.04; Table 4). The full model that used IQ, sex, executive functioning, and the subscales of the AQ was able to predict CAT-Q scores (Model 2) with statistical significance, F(8, 238) = 15.088, p < 0.001; R² = 0.336.

Table 4.

Hierarchical Regression Analysis Predicting Camouflaging Using Autistic Traits

Variable	CAT-Q
	Model 1		Model 2
	B	β	B	β
Constant	64.327		54.835
IQ	−0.154	−0.056	−0.082	−0.030
Sex	−1.132	−0.022	0.537	0.010
BRIEF-MI	0.790**	0.334	0.382*	0.162
AQ social skills			1.471*	0.144
AQ attention switching			1.588*	0.148
AQ attention to details			0.764	0.075
AQ communication skills			3.597**	0.302
AQ imagination			−0.105	−0.007
R 2	0.114		0.336
F	10.429**		15.088**
ΔR2	—		0.222
ΔF	—		15.957**

n = 247.

^* p < 0.05.

^** p < 0.001.

Given that both social competence and autistic traits were found to account for significant variance in camouflaging behaviors, we conducted a third regression to determine whether the total MSCS score could account for additional variance over and above the total AQ score, IQ, sex, and executive functioning (Table 5). We started with Model 1 (see above). Self-report of total autism characteristics scores were then added to the prediction of CAT-Q scores (Model 2), leading to a statistically significant increase in R² of 0.192, F(4, 242) = 26.672, p < 0.001, with higher AQ scores associated with increased camouflaging. Finally, total MSCS scores were added to the prediction of CAT-Q scores (Model 3), leading to another significant increase in R² of 0.024, F(5, 241) = 23.736, p < 0.001, with lower MSCS scores associated with increased camouflaging.

Discussion

The aim of the current study was to explore the ability of social competence to predict camouflaging behaviors in a nonautistic sample. We found that after controlling for IQ, sex, and executive functioning levels (BRIEF-MI), social competence accounted for additional variance (25%) in camouflaging behaviors, with lower social competence associated with increased camouflaging behaviors. When considering the subscales of the MSCS, the social motivation, social inferencing, social knowledge, and nonverbal sending skills subscales all significantly contributed, and the emotional regulation subscale approached a significant contribution, to the model. Thus, these findings indicate that social competence accounts for significant variance in CAT-Q scores, suggesting that lower social competencies are sufficient, at least in nonautistic adults, to predict camouflaging behaviors.

In addition to looking at the relationship between social competency and camouflaging behaviors, we also aimed to replicate previous findings that report that autistic characteristics were associated with camouflaging behaviors in a nonautistic sample.⁸ Similar to the previous findings, we found that after controlling for IQ, sex, and executive functioning levels (BRIEF-MI), self-reported autism characteristics accounted for additional variance (22%) in camouflaging behaviors, with more autistic characteristics associated with increased camouflaging behaviors. Interestingly, when considering the subscales of the AQ, only the social skills, attention switching/tolerance to change, and communication skills subscales contributed significantly to the model. These results suggest that it is largely the social aspects of autism, as measured by the AQ, that are related to camouflaging.

Given that social competence and autism characteristics were highly correlated, it is possible that both the MSCS and AQ were measuring the same variance in camouflaging behaviors. To investigate whether social competence alone could account for unique variance in camouflaging behaviors, over and above that accounted for by autism characteristics, an additional regression was conducted where both AQ and MSCS scores were included in the model. The results indicated that social competence accounted for additional variance (2.4%) in camouflaging behaviors, over and above autism characteristics. This decrease in the variance accounted for with AQ scores included in the model suggests that the MSCS and AQ measure overlapping constructs. This is not surprising given that social difficulties are a defining characteristic of autism, and the AQ measures autistic traits. However, given the MSCS accounted for a small, but significant, amount of additional variance suggests that social competence alone can account for a significant variance in camouflaging behaviors.

Together with the results of previous studies, these results suggest that both autistic individuals^6–10 and nonautistic individuals use camouflaging techniques to “fit-in” or mask the undesirable aspects of their social behavior when in social situations. Given that social difficulties are a defining feature of the ASD diagnosis, it is likely that camouflaging is more prevalent among autistic individuals, relative to nonautistic individuals. However, the current results replicate previous findings⁸ and suggest that the camouflaging response to social difficulties does not appear to be unique to autistic individuals. However, it is important to study the differences in camouflaging across autistic and nonautistic samples so as to understand the qualitative aspects of the phenomena.

It is important to acknowledge that whereas poor social skills are associated with increased camouflaging behaviors, these camouflaging behaviors are performed to comply with an external, or environmental, demand to behave in a particular manner.⁶ This highlights the importance of conceptualizing the social challenges that autistic people face within a broader relational context of interactions that include bidirectional influences and not simply a matter of reducing the stigma associated with the person's autistic traits.^5,27 Avoiding autism stigma is not an easy feat, as the stigma surrounding autism is present in peers by early childhood²⁸ and continues into adolescence and adulthood.⁵ However, recent research has demonstrated that brief training geared toward increasing nonautistic individuals' understanding of autism can influence their overall acceptance of their autistic peers.²⁹ Given the strong association between camouflaging and social competence in nonautistic individuals, this finding warrants more research to investigate what conditions elicit camouflaging behavior in all individuals with low social competence and what is specific to autistic people. This will contribute to the design of better social sensitivity training for all individuals, to reduce the negative perceptions of atypical social behavior without putting the onus solely on individuals to comply with social conventions.

Given camouflaging behaviors were investigated using a standardized self-report measure with nonautistic participants, we are unable to qualify how camouflaging behaviors differ across autistic and nonautistic individuals. We are also not able to conclude whether the effects of these camouflaging behaviors are the same across autistic and nonautistic individuals without the inclusion of a comparison group of autistic participants. Previous self-report evidence suggests that there may be differences in the ways in which autistic and nonautistic individuals camouflage that result in differences in the negative psychological effects of these behaviors.^6,10,18 Future research comparing the similarity and differences in camouflaging across these populations is necessary. Given that the current study comprised a sample of university students, the generalizability of this study is limited. The sample is homogenous in age and may face social challenges or demands that nonstudents may not encounter, which could influence the extent to which they use camouflaging behaviors. Future research should include a sample that is more representative of the greater population. In addition, given that social competence and autistic traits were also investigated using self-report measures, it is possible that the responses of some participants were influenced by a social desirability bias. Interestingly, social desirability bias may be conceptualized as a camouflaging strategy: reporting fewer autistic traits and better social competence to appear more socially desirable. However, given that we did not include a measure of social desirability bias, it is unclear what impact this may have had on our data. Finally, given the strong correlation between AQ and MSCS scores, this limited our ability to compare the amount of independent variance explained by the two measures.

The results of this study replicate previous findings that nonautistic adults also adopt camouflaging strategies to compensate for their social difficulties.⁸ Furthermore, these results suggest that camouflaging strategies are strongly associated with both measures of social competence and autistic characteristics, suggesting that poor social competencies predict camouflaging behaviors in nonautistic individuals. Since camouflaging appears to be a strategy that is adopted by both autistic and nonautistic individuals, the findings have implications for how we conceptualize camouflaging and design prevention/intervention techniques. Rather than see camouflaging as a phenomenon that occurs exclusively in autistic people because of their disability, we can view it as a response that both autistic and nonautistic people may use when they perceive themselves to lack the necessary social competencies that are expected of them within their social contexts. Because both autistic traits and social competency are related to camouflaging behaviors, we can begin to explore which characteristics are more likely to evoke camouflaging in autistic individuals and how this may be similar or different in nonautistic individuals. This, in turn, can provide more precise hypotheses on which conditions are more likely to evoke camouflaging in autistic people and what is shared with or unique to nonautistic people. This more nuanced understanding will ultimately contribute to the development of more tailored approaches to prevent and/or reduce the negative impact of camouflaging behaviors.

Authorship Confirmation Statement

N.S., H.A., T.B., and G.I. conceived and designed the experiment. N.S. and T.B. collected the data. N.S. and T.B. analyzed the data. N.S. and G.I. prepared the initial draft of the article, while H.A. and T.B. reviewed and edited the final draft. All authors reviewed and approved the final draft before submission. The article has been submitted solely to this journal and is not published, in press, or submitted elsewhere.

Author Disclosure Statement

All authors declare that they have no actual or potential competing financial interests.

Funding Information

Natural Science and Engineering Research Council of Canada (NSERC) grant #408052 to G.I. and NSERC post-doctoral fellowship #502976 granted to N.E.S.