Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Jun 5;56(12):4940–4951. doi: 10.1007/s10803-025-06893-x

Reliability and Validity of the Chinese Version of the Social Skills Improvement System Autism Spectrum Scale

Manying Hsieh 1,2, Hui-Ting Wang 3,4,✉, Yu-ping Chen 3,4, Wei-cheng Wang 5,6
PMCID: PMC13694331  PMID: 40471399

Abstract

The purpose of this study was to examine the reliability and validity of the Chinese version of the SSIS Autism Spectrum Scale Teacher Form, a scale of the SSIS-Rating Scales (Social Skills Improvement System Rating Scales) created by Gresham and Elliott. The scale consists of 15 items divided into two subscales: social skills and problem behaviors related to autism. The participants included 541 school-age children aged 4 to 18 years old (266 females, 275 males). The Autism Behavior Checklist-Taiwan Version Teacher Form was used as a criterion measure. Data analyses were performed using SPSS 25 and Mplus 8.3 statistical package programs, with procedures including reliability testing, criterion-related validity testing, and confirmatory factor analysis (CFA). The results suggested that the SSIS Autism Spectrum Scale Teacher Form has adequate internal consistency reliability, test-retest reliability, and convergent validity. In addition, the confirmatory factor analysis results supported the two-factor structure of the Chinese version of the SSIS Autism Spectrum Scale Teacher Form.

Keywords: Autism, Confirmatory factor analysis, Construct validity, Reliability

Autism Spectrum Disorder (ASD) is a neuro-developmental disorder characterized by repetitive behaviors, limited interests, and challenges with social interaction (Hyman et al., 2020). The severity of symptoms can vary greatly among individuals and can seriously impede the typical social functions of children. As a result, most children with autistic traits face significant challenges during early school years, stemming from specific autism-related behavioral problems and social skill deficits. These challenges often make their early school life exceptionally difficult, hindering their full participation in the classroom and preventing them from fully realizing their potential. In recent decades, the prevalence of autism increased in various countries, including the United States, South Korea, and Taiwan, partially due to changes in diagnostic criteria and improved screening tools (Fombonne, 2018; Zeidan et al., 2022). The change of the diagnostic criteria in the DSM-5 led to the inclusion of milder forms of autism, resulting in a larger number of identified cases (Davidovitch et al., 2020; Hoang et al., 2019). Additionally, the availability and accessibility of screening tools and diagnostic instruments has improved in past decades, also making it possible for even children with milder forms of autism to potentially receive a diagnosis (Zeidan et al., 2022). Psychologists have recently shown increasing interest in the social skills learning needs of school-aged children, especially those with autistic traits. With more children being diagnosed with ASD, there is an urgent need for more tools and strategies to address this issue (Conroy et al., 2007).

The Social Skills Improvement System Autism Spectrum Scale is a standardized assessment tool and is also a part of the well-known Social Skills Improvement System Rating Scales (SSIS-RS) created by Gresham and Elliott (2008). Its purpose is to assist in the development and evaluation of social skills interventions for children, rather than to serve as a diagnostic tool. It has two versions, parent and teacher forms, and assesses two key dimensions, the social skills domain and the problem behaviors domain related to autism. The social skills domain assesses skills such as appropriate responses, turn-taking, and maintaining eye contact during communication, while the problem behavior domain evaluates symptoms such as stereotyped behavior, social withdrawal, resistance to change, and difficulty with transitions. In addition, by reversing the score of the social skills domain related to autism and adding it to the score of behavioral problems, a total score can be obtained. The higher the total score on the scale, the more social skills difficulties related to autism are indicated, according to the recommendations of the SSIS-RS manual (Gresham & Elliott, 2008).

Development of the SSIS Autism Spectrum Scale and Related Research

SSIS Autism Spectrum Scale is developed based on Gresham and Elliott’s theory of social skills deficits (2008) and is one of a series of scales in the SSIS-RS, an updated scoring scale derived from the original Social Skills Rating System (SSRS; Gresham & Elliott, 1989). The SSRS demonstrated solid psychometric qualities, including internal consistency, construct validity, convergent validity, and criterion validity in a small sample of preschoolers with autism (Wang et al., 2011). These properties indicate that the SSRS is a reliable tool for assessing social skills in young children with ASD in natural settings. However, it may not be highly effective for tracking social skills development over time or for evaluating the outcomes of interventions in these children.

The SSIS Autism Scale was then included in the SSIS-RS in 2008, with its items based on the fourth edition of the Diagnostic and Statistical Manual (DSM-IV). According to Gresham and Elliott’s social skills deficit theory (2008), there is a competitive relationship between the two constructs of the SSIS Autism Spectrum Scale. In other words, children with higher autism-related social skills tend to have fewer autism-related behavioral problems, and vice versa. This implies that improving social skills in children with autism could potentially reduce their behavioral problems, and that interventions and educational programs focused on enhancing social skills are crucial. Although more longitudinal studies are needed to test this hypothesis, undoubtedly, the theoretical perspective of Gresham and Elliott (2008) provides a more comprehensive picture for interventions in children with autism. The SSIS Autism Spectrum Scale is a unique assessment tool in its emphasis on school settings. It helps educators and psychologists identify areas needing improvement in school-aged children with ASD. The teaching units in the SSIS Manual align with the factors and items measured by the tool, ensuring the assessment’s relevance to teaching practice. Its connection to educational practice makes it a valuable resource for school-aged children with ASD, and therefore recommended by child psychologists and widely used (Panayiotou et al., 2022).

Nonetheless, the SSIS Autism Spectrum Scale has some weaknesses. Since its creation, there have been few studies evaluating its validity. Although the developers included it as a potential screening tool within the SSIS, aiming to differentiate children with autism, so far there is little information and few relevant empirical studies that examine its usefulness for autism screening (Crosby, 2011). Moreover, there is limited evidence supporting the structural validity of the SSIS Autism Spectrum Scale. Although the developers integrated theory, empirical research, and the original factor framework of the SSRS, and conducted various statistical analyses on the initial items, most SSIS-RS subscales exhibited poor fit in their original factor models in the confirmatory factor analysis (CFA) (Panayiotou et al., 2022). Nader et al. (2023) conducted an exploratory factor analysis (EFA) of the SSIS Autism Spectrum Scale on children aged 4 to 12 years old with ASD in the United States. The study found that while items in the social skills domain of the scale were correctly loaded, six out of the seven items in the problem behavior domains failed to load correctly, indicating that the SSIS Autism Spectrum Scale has insufficient construct validity. This result may be because all the children in the sample were diagnosed with Autism Spectrum Disorder (ASD), which could influence the assumptions underlying EFA and its analysis results. In addition, Cheung et al. (2017) conducted a study in Hong Kong to examine the reliability and validity of the translated unpublished Chinese version of the SSIS Autism Spectrum Scale. The research found that the SSIS Autism Spectrum Scale demonstrated good internal consistency reliability, with a Cronbach’s α value of 0.83, and effectively distinguished between children with typical development and those with Autism Spectrum Disorder (ASD). Cheung et al.‘s study showed that the Autism Spectrum Scale is very sensitive in identifying autism-related behaviors in children. The study also found that the autism group scored highest on the Autism Spectrum Scale in various behavior problems. However, these results should be interpreted cautiously due to the small sample size, and the study did not investigate the support needs of the autism group. Overall, the inconsistent results in existing research on the SSIS Autism Spectrum Scale highlight the need for further validation and a more thorough examination of its validity and reliability, in order to better support children with ASD.

Factors That May Affect Social Skills

Furthermore, the appropriateness of social skills is not universal or independent of context, but rather, it is context-dependent (Friedberg, 2012). The social skills of children with ASD may be influenced by some contextual factors such as gender and culture. Therefore, further research is needed to provide evidence of the reliability and validity of the SSIS Autism Spectrum Scale across genders and across multiple cultures.

Studies have found that boys were diagnosed with ASD at a rate approximately four times higher than girls (Beggiato et al., 2017; Zeidan et al., 2022). Boys with ASD tended to experience more difficulties in early social development than girls with ASD, showing a higher frequency of restricted repetitive behaviors (Lai et al., 2015; Supekar & Menon, 2015), less motivation for social contact (Sedgewick et al., 2016), poorer social skills and fewer friendships (Head et al., 2014). In addition, Wood-Downie et al. (2021) conducted a meta-analysis and found that most studies reported females with ASD had better social interactions and skills compared to their male peers, and this trend was also observed in non-autistic individuals.

In addition, social skills refer to the ability of an individual to appropriately interact with others in various situations (Olmeda & Kauffman, 2003), and their effectiveness is determined in various situations (Nangle et al., 2010). They involve understanding and evaluating social situations based on relevant social rules, and this comprehension is critical for developing appropriate social behaviors, implying that different social environments may prioritize and emphasize different sets of skills. Previous research suggested that social-cultural contexts may play an important role influencing the development of social skills (Friedberg, 2012). For instance, Hsu (2007) conducted a cross-cultural study to compare the communication skills between Americans and Taiwanese. Hsu (2007) found that Taiwanese people had a lower self-perceived communication capacity, were less willing to communicate with others, exhibited more shyness, but showed higher levels of self-disclosure compared to Americans.

Previous research indicated that social skills training programs for children with ASD can improve facial emotion recognition, reduce autism traits, behavior problems, social anxiety, and loneliness (Afsharnejad et al., 2021), and enhance social knowledge and general social skills (Marro et al., 2019). A meta-analysis by Rao et al. (2008) found that only one out of ten social skills training studies for youth with Asperger’s syndrome or high-functioning autism reported favorable outcomes across all measured aspects. Thus, Rao’s study highlighted the need to further develop and improve assessment tools to accurately measure the progress of children with ASD in social skills. Considering that effective intervention measures and reliable evaluation methods are crucial for supporting the social learning needs of individuals with ASD, more research is needed on tools for assessing the performance of social skills related to autistic traits.

The purpose of this study was to examine the reliability and validity of the Chinese version of SSIS Autism Spectrum Scale Teacher Frm (Hsieh & Wang, 2025). This study only covers the teacher’s form of the scale and explores whether the theoretical model of the scale developed in Western cultures is also applicable to other cultures and contexts. Although traditional Taiwanese culture emphasizes interpersonal etiquette and social roles, western values have influenced today’s Taiwanese society, making it more open and respectful of individual personality. Thus, the emphasis on interpersonal etiquette is not as obvious as it once was. In addition, good assessment tools can help determine whether an intervention has achieved its intended goals and identify areas that need further modification. It is believed that well-designed and rigorously tested assessments may ultimately assist in the development and planning of programs for children with ASD and aid in these children’s future life adaptation.

Methods

Procedure and Participants

The study used a convenience sample, recruiting participants from 17 counties and cities in Taiwan through public schools in 2020. The schools are selected based on the population proportion of the four administrative districts in Taiwan: eastern, western, southern, and northern areas. This selection process ensured that the chosen schools reflected diverse demographic characteristics and educational environments. Teachers in these schools were invited to participate, and each teacher rated only 1–3 typically developing students using anonymous questionnaires, provided they had known the student for at least a year. The typically developing students were randomly selected based on the last digit of their student ID number being 5. However, considering the limited number of students with special education status or ASD in most schools, teachers were allowed to rate multiple students with special education status or ASD to ensure these students were not neglected. Overall, out of the 711 questionnaires sent to schools, 541 were completed and returned to the researchers, resulting in a response rate of approximately 76%.

Among the final sample of 541 school-age children and adolescents, 266 were female and 275 were male, with a mean age of 11.77 years (SD = 3.84). The ages of the participants were evenly distributed: 2.9% were under 6 years old, 7.4% were 6 years old, 7.4% were 7 years old, 7.4% were 8 years old, 7.6% were 9 years old, 7.8% were 10 years old, 7.4% were 11 years old, 7.6% were 12 years old, 7.8% were 13 years old, 7.8% were 14 years old, 7.6% were 15 years old, 7.6% were 16 years old, 7.4% were 17 years old, and 6.8% were 18 years old. Among the participants, 6.4% were identified as Taiwanese aborigines, 4.1% were from immigrant families, 22.5% were qualified for special education services, and 10.0% were diagnosed by psychiatrists as having ASD (See Table 1).

Table 1.

The characteristics of the sample

Percentage N
Special Education Status 21.10% 114
Autism 10.00% 54
Immigrant Family Background 4.10% 22
Aboriginal Person 5.90% 32
Gender Male 50.80% 275
Female 49.20% 266
Region Western 17.80% 96
Northern 48.10% 260
Eastern 9.80% 53
Southern 24.30% 131
Missing 1
Open in a new tab

Instruments

Autism Behavior Checklist-Taiwan Version Teacher Form. The Taiwanese version of the Autism Behavior Checklist (ABC-T; Huang & Wu, 2013) is a standardized questionnaire consisting of 47 items that evaluates symptoms of autism. It’s adapted from the well-known Autism Behavior Checklist of the Autism Screening Instrument for Education Planning (ABC-3rd edition; Krug et al., 1980). The questionnaire contains five factors: Sensation and Perception, Relation and Connection, Body and Object Use, Language and Social and Self-Help. Each item is rated with a “yes” or “no”, and higher scores indicate more severe symptoms of ASD. Previous research showed that the ABC-T had good internal consistency (Kuder-Richardson reliability coefficients: 0.96 for the whole ABC-T scale; 75 for the Sensation and Perception subscale, 0.86 for the Relation and Connection subscale, 0.84 for the Body and Object Use subscale, 0.82 for the Language subscale, and 0.78 for the Social and Self-Help subscale), satisfactory test-retest reliability (r =.89), and appropriate convergent and discriminant validity (Chang & Wu, 2006; Huang & Wu, 2013). In the present study, the ABC-T also demonstrated good internal consistency (Kuder-Richardson reliability coefficients: 0.92 for the whole ABC-T scale; 0.77 for the Sensation and Perception subscale, 0.86 for the Relation and Connection subscale, 0.76 for the Body and Object Use subscale, 0.80 for the Language subscale, and 0.70 for the Social and Self-Help subscale).

SSIS Autism Spectrum Scale Teacher Form. The SSIS Autism Spectrum Scale Teacher Form (Gresham, & Elliott, 2008) contains 15 items, which can be divided into two subscales: the social skills domain and the problem behaviors domain related to autism. The social skills domain assesses autism-related social skills, with higher scores indicating better social skills. The problem behaviors domain measures the severity of behavioral issues in ASD, with higher scores implying more severe autism-related behavioral problems. Items are rated on a 4-point Likert scale ranging from 0 (“not true”) to 3 (“very true”). The current study used the Chinese version of the SSIS Autism Spectrum Scale Teacher Form (Hsieh & Wang, 2025), published by the authors in Taiwan this year. A translator who lived in the United States and was fluent in both Chinese and English translated the items of the scale into Chinese, and another translator back-translated the Chinese version into English. Two university professors, specializing in English and child development, reviewed both the original and back-translated versions of the scale and found they were highly consistent. The Chinese version of the SSIS Autism Spectrum Scale Teacher Form showed adequate internal consistency, with an alpha value of 0.90 for the social skills domain related to autism and 0.88 for the problem behaviors domain related to autism.

Data Analyses

Data was analyzed using SPSS 25 and Mplus 8.3 statistics programs, with procedures including internal consistency reliability testing, test-retest reliability testing, convergent validity testing, and construct validity testing.

First, internal consistency reliability refers to the consistency of results across items (Mohan, 2023), and Cronbach’s alpha is the most widely used method for evaluating the internal consistency reliability of a scale. A commonly accepted guideline is that a Cronbach’s Alpha between 0.7 signifies an acceptable level of reliability, while a value of 0.8 or higher indicates a very good level of reliability. Second, test-retest reliability refers to the extent to which scores remain consistent over time (Mohan, 2023). This can be assessed by administering the same test to the same group at two different times and calculating the Pearson correlation coefficients between the scores. Generally, Cronbach’s alpha coefficients of 0.7 or above demonstrate acceptable reliability, while coefficients of 0.8 or higher indicate very good reliability (Taber, 2018). Third, convergent validity refers to the extent to which a scale is closely related to other established measures that assess the same or similar constructs. To evaluate convergent validity, a sample from the target population completes both the new scale and established measures simultaneously. The Pearson correlation coefficients between the new scale and the related measures are then computed. Generally, significant high correlations indicate good convergent validity. Fourth, construct validity refers to how accurately a scale measures its theoretical concept. This can be assessed by performing CFA. Good model fit and adequate factor loadings provide evidence supporting the construct validity of a scale. In the current study, the Chi-square, Root Mean Square Error of Approximation (RMSEA), and Comparative Fit Index (CFI) were calculated and compared to evaluate how well the hypothesized measurement model of the SSIS Autism Spectrum Scale Teacher Form fits the observed data. Generally, a good fit is indicated by a low Chi-square statistic, an RMSEA of less than 0.05, and a CFI greater than 0.90 (Chen, 2007; Hair et al., 2018). Moreover, to evaluate measurement invariance, which refers to the psychometric equivalence of an instrument across diverse social groups (Putnick & Bornstein, 2016), we conducted Multi-Group Confirmatory Factor Analysis (MGCFA) using the Mplus statistics program with the Weight Least Square with Mean and Variance (WLSMV) estimation, which is a robust method to analyze ordered categorical data. This process involved computing CFA with data from multiple groups simultaneously. In the current study, measurement invariance was assessed using changes in Chi-Square statistics, CFI, and RMSEA fit indices.

Results

Given that the SSIS Autism Spectrum Rating Scales Teacher Form has not been tested in the Taiwanese population, reliability and validity analyses were conducted. Reliability was assessed using Cronbach’s α and test-retest correlation, while validity was evaluated through convergent validity tests and CFA. The characteristics of the sample and the psychometric properties of the SSIS Autism Spectrum Rating Scales Teacher Form are reported in Tables 1 and 2. Considering that the distribution of child behavioral problems tends to be slightly skewed, normality tests were then conducted using skewness and kurtosis measures. Skewness measures the level of asymmetry of the distribution. Positive skewness indicates a distribution with a longer right tail and most values left of the mean, while negative skewness indicates a longer left tail with most values right of the mean. Kline (2023) suggested a criterion for a substantial departure from normality as an absolute skew value greater than 3. In this study, the absolute skewness values for all items and the summed scores of the two constructs in the SSIS Autism Spectrum Scale were below 3 (social skills domains related to autism: -0.29; problem behaviors domains related to autism: 0.70). Kurtosis measures the sharpness of a distribution’s peak, with a value above 7 indicating a significant deviation from normality (Kline, 2023). In this study, the absolute kurtosis values for all items and the summed scores of the two constructs were below 7 (social skills domains related to autism: -0.22; problem behaviors domains related to autism: -0.06), indicating no significant deviation from normality.

Table 2.

The psychometric properties of the SSIS autism spectrum scale

M SD Skewness Kurtosis Alpha Test-retest Reliability
Q10 2.02 0.74 -0.3 -0.41 0.89 0.80
Q19 2.07 0.77 -0.36 -0.6
Q20 2.02 0.74 -0.37 -0.19
Q29 1.87 0.76 -0.27 -0.28
Q30 2.1 0.73 -0.51 0.08
Q38 1.98 0.78 -0.3 -0.5
Q39 1.92 0.75 -0.23 -0.38
Q40 1.92 0.73 -0.25 -0.23
Sum 15.9 4.63 -0.29 -0.22
Q48 1.73 0.85 -0.04 -0.75 0.86 0.85
Q50 1.05 0.79 0.47 -0.12
Q54 0.76 0.82 0.86 0.02
Q56 0.66 0.79 1.02 0.42
Q60 0.73 0.82 0.89 -0.02
Q66 0.66 0.82 1.1 0.52
Q72 0.73 0.83 0.88 -0.06
Sum 6.32 4.32 0.7 -0.06
Total 14.42 8.06 0.47 -0.18
Open in a new tab

Reliability

The internal consistency analysis of the SSIS Autism Spectrum Scale demonstrated good overall reliability, with Cronbach’s alpha coefficients of 0.89 for the social skills domains related to autism and 0.86 for the problem behaviors domains related to autism. Additionally, a correlation coefficient of − 0.60 was found between social skills and problem behaviors domains related to autism, suggesting a significant but negative relationship between the two constructs.

This study used a convenience sample for test-retest reliability. Thirty participants were sufficient, so they weren’t randomly selected nationwide nor representative of the population under study. Instead, they were the target population of the SSIS Autism Spectrum Scale, including 13 boys and 17 girls (Mean Age = 9.93 years, SD = 2.68 years), all attending the same public junior high in southern Taiwan. Choosing participants from a single school also made it easier to conduct the follow-up survey in the research process. The SSIS Autism Spectrum Scale showed adequate test-retest reliability over a two-week interval, with coefficients of 0.80 for the social skills domain related to autism and 0.85 for the problem behaviors domain related to autism. Overall, these findings suggested that the SSIS Autism Spectrum Scale has adequate internal consistency and stability over time.

Convergent Validity

This study assessed the convergent validity of the SSIS Autism Spectrum Scale using a convenience sample of 29 participants (23 boys and 6 girls, Mean Age = 11.72 years, SD = 3.33 years). The participants were not randomly selected nationwide nor representative of the population under study, but they were the target population of the SSIS Autism Spectrum Scale and were included in the final sample. The ABC-T Teacher Form (Huang & Wu, 2013) was used as the criterion measure. Pearson correlation coefficients between the criterion measure and the SSIS Autism Spectrum Scale were then computed.

Results showed that the SSIS Autism Spectrum Scale Teacher Form was significantly related to the ABC-T Teacher Form. Specifically, the social skills domain of the SSIS Autism Spectrum Scale showed significant negative correlations with the Relation and Connection subscale, Language subscale and Sensation and Perception subscale of ABC-T (r = −.73 for Relation and Connection, r = −.51 for Language, and r = −.56 for Sensation and Perception; p <.01); while the problem behaviors domain of the SSIS Autism Spectrum Scale showed significant positive correlations with all five subscales of ABC-T (r =.83 for the Sensation and Perception subscale, r =.84 for the Relation and Connection subscale, r =.55 for the Body and Object Use subscale, r =.71 for the Language subscale, and r =.57 for the Social and Self-Help subscale; p <.01). These results suggested that the SSIS Autism Spectrum Scale Teacher Form has adequate convergent validity.

Confirmatory Factor Analysis

Descriptive statistics for each item of the SSIS Autism Spectrum Scale Teacher Form were reported separately for boys and girls (see Table 2). Table 3 presents the means, standard deviations, and bivariate correlations by gender. Because a normality test was conducted to determine whether the data was normally distributed, and the results indicated that the normality assumptions were not violated, the data was suitable for CFA (Kline, 2023).

Table 3.

Means, standard deviations, and bivariate correlations for girls and boys (N = 541)

Boys
Q10 Q19 Q20 Q29 Q30 Q38 Q39 Q40 Q48 Q50 Q54 Q56 Q60 Q66 Q72 M SD
Q10 - 0.64 0.58 0.47 0.57 0.57 0.45 0.53 − 0.29 − 0.34 − 0.42 − 0.36 − 0.46 − 0.40 − 0.41 1.86 0.73
Q19 0.61 - 0.55 0.52 0.59 0.64 0.58 0.57 − 0.33 − 0.42 − 0.46 − 0.44 − 0.48 − 0.51 − 0.48 1.89 0.82
Q20 0.60 0.60 - 0.39 0.47 0.48 0.33 0.44 − 0.32 − 0.37 − 0.43 − 0.29 − 0.42 − 0.41 − 0.43 1.87 0.75
Q29 0.45 0.49 0.39 - 0.58 0.53 0.57 0.49 − 0.19 − 0.31 − 0.32 − 0.37 − 0.37 − 0.36 − 0.26 1.69 0.77
Q30 0.51 0.47 0.54 0.40 - 0.57 0.54 0.53 − 0.23 − 0.34 − 0.41 − 0.44 − 0.42 − 0.42 − 0.32 1.96 0.75
Q38 0.59 0.58 0.50 0.51 0.45 - 0.56 0.58 − 0.22 − 0.40 − 0.35 − 0.36 − 0.39 − 0.41 − 0.37 1.75 0.78
Q39 0.52 0.58 0.42 0.49 0.40 0.62 - 0.58 − 0.14 − 0.36 − 0.41 − 0.48 − 0.37 − 0.42 − 0.29 1.80 0.73
Q40 0.54 0.48 0.45 0.45 0.43 0.55 0.60 - − 0.19 − 0.34 − 0.36 − 0.39 − 0.35 − 0.38 − 0.35 1.77 0.72
Q48 − 0.31 − 0.28 − 0.28 − 0.24 − 0.14 − 0.26 − 0.16 − 0.18 - 0.39 0.39 0.25 0.44 0.34 0.36 1.89 0.83
Q50 − 0.40 − 0.32 − 0.33 − 0.25 − 0.21 − 0.36 − 0.28 − 0.26 0.42 - 0.54 0.43 0.51 0.48 0.44 1.18 0.77
Q54 − 0.37 − 0.40 − 0.33 − 0.34 − 0.26 − 0.40 − 0.33 − 0.30 0.42 0.52 - 0.54 0.64 0.60 0.60 0.93 0.89
Q56 − 0.42 − 0.41 − 0.33 − 0.38 − 0.21 − 0.41 − 0.41 − 0.31 0.34 0.47 0.58 - 0.58 0.47 0.38 0.72 0.81
Q60 − 0.29 − 0.34 − 0.29 − 0.21 − 0.13 − 0.40 − 0.30 − 0.21 0.36 0.55 0.63 0.51 - 0.65 0.58 0.88 0.88
Q66 − 0.42 − 0.32 − 0.34 − 0.24 − 0.28 − 0.30 − 0.25 − 0.27 0.37 0.50 0.53 0.49 0.57 - 0.65 0.86 0.91
Q72 − 0.33 − 0.31 − 0.26 − 0.25 − 0.28 − 0.38 − 0.21 − 0.22 0.34 0.45 0.48 0.43 0.51 0.55 - 0.92 0.89
Girls
M 2.18 2.25 2.18 2.06 2.25 2.22 2.03 2.07 1.56 0.92 0.58 0.60 0.57 0.46 0.52
SD 0.72 0.68 0.69 0.71 0.67 0.70 0.75 0.70 0.83 0.79 0.70 0.75 0.73 0.66 0.70
Open in a new tab

Note. All correlation coefficients were statistically significant (p <.05)

Measurement invariance—a statistical property where a measure assesses a latent construct in the same way across different groups—was evaluated (Chen, 2007; Finch, 2014). All models in the current study were analyzed using CFA with the Weighted Least Squares Mean and Variance (WLSMV) estimator to test the hypothesized 2-factor model of the SSIS Autism Spectrum Scale for two groups: boys and girls (see Table 4). Typically, CFI and TLI values greater than 0.95 and RMSEA values smaller than 0.06 indicate relatively excellent model–data fit (Hu & Bentler, 1999). Firstly, the model was run with all parameters freely estimated across gender. The results revealed that this model (an unconstrained model, namely Model 1) had a good fit to the data (χ2(178) = 344.936, p <.01, CFI = 0.984, TLI = 0.981, RMSEA = 0.059, N = 541) and all standardized factor loadings were significant and above 0.50. Secondly, the model was rerun with all factor loadings of the items constrained to be equal across gender. The results showed that this model, called Model 2, had a satisfactory fit to the data (see Table 4). A chi-square difference test was then conducted to compare the fits between Models 1 and 2. No significant difference was found between the models (Δχ2(13) = 19.011, p >.05). Moreover, the CFI value did not change by more than 0.01, and the RMSEA value did not change by more than 0.015. Based on Chen’s (2007) guidelines, since there were no significant differences between the two models and the CFI and RMSEA values changed very little, Model 2 fitted the empirical data as well as Model 1, which supported the hypothesis of metric equivalence. Considering that Model 2 fitted the data as well as Model 1 and was more parsimonious, we selected Model 2 as the preferable model. Thirdly, we set up Model 3, which constrained all factor loading and means of the items to be equal across gender, and reran it. The results showed that Model 3 fit to the data well (see Table 4). The fit of Models 2 and 3 was then compared using a chi-square difference test (see Table 4), and no significant difference was found between the models (Δχ2(28) = 40.552, p >.05). Additionally, the CFI value did not increase or decrease by more than 0.01, and the RMSEA value did not increase or decrease by more than 0.015. Based on Chen’s (2007) guidelines, since there were no significant differences between the two models and the CFI and RMSEA values changed very little, Model 3 fitted the empirical data as well as Model 2, which supported the hypothesis of metric equivalence. Considering that Model 3 fitted the data as well as Model 2 and was most parsimonious, Model 3 was then selected as the chosen model (see Table 4; Fig. 1).

Table 4.

Tested models (N = 541)

χ2 df p CFI TLI RMSEA Difference test
Δdf Δχ2 p ΔCFI ΔRMSEA
Model 1 (Unconstrained) 344.936 178 0.00 0.984 0.981 0.059
Model 2 (Metric Equivalence) 363.947 191 0.00 0.983 0.981 0.058 13 19.011 0.123 -0.001 0.000
Model 3 (Scalar Equivalence) 404.499 219 0.00 0.982 0.983 0.056 28 40.552 0.059 -0.001 0.002
Open in a new tab

Fig. 1.

Fig. 1

Open in a new tab

The confirmatory factor analysis model for the SSIS autism spectrum scale (N = 541). Note. Standardized path coefficients were shown in the figure, with boys listed first and girls listed in parentheses. All factor loadings were statistically significant (p <.01)

Overall, the final model, Model 3, confirmed a two-factor theoretical framework of the SSIS Autism Spectrum Scale with a good fit to the model (χ²( 219) = 404.499, p <.01, CFI = 0.982, TLI = 0.983, RMSEA = 0.056, N = 541). All standardized factor loadings were reported separately for boys and girls and were all above 0.50 (see Table 5). In this final model, the latent correlation between the two domains was found to be negatively significant, r = -.73 for boys and r = -.63 for girls.

Table 5.

Factor loadings for boys and girls for the SSIS autism spectrum scale (N = 541)

Items Unstandardized Factor Loadings Standardized Factor
Loadings
Boys Girls Boys Girls
Social Skills Related to Autism
Q10 (Responses in Conversations) 1.00 1.00 0.83 0.85
Q19 (Good Interaction with Others) 1.27 1.27 0.88 0.86
Q20 (Taking Turns) 0.72 0.72 0.73 0.79
Q29 (Joining in Activities) 0.73 0.73 0.73 0.70
Q30 (Eye Contact) 0.92 0.92 0.80 0.70
Q38 (Concern for Others) 0.95 0.95 0.81 0.86
Q39 (Initiating Conversations) 0.87 0.87 0.79 0.78
Q40 (Using Gestures) 0.85 0.85 0.78 0.75
Problem Behaviors Related to Autism
Q48 (Unusual Preoccupations) 1.00 1.00 0.55 0.59
Q50 (Response to Routine Change) 1.62 1.62 0.73 0.74
Q54 (Stereotypic Behaviors) 2.33 2.33 0.84 0.87
Q56 (Social Withdrawal) 1.83 1.83 0.77 0.77
Q60 (Repetitive Behaviors) 2.62 2.62 0.86 0.83
Q66 (Nonfunctional Routines) 2.45 2.45 0.85 0.84
Q72 (Odd Physical Gestures) 1.91 1.91 0.78 0.76
Open in a new tab

Note. All of the factor loadings were significant (p <.05)

Discussion

This study examined the reliability and validity of the SSIS Autism Spectrum Scale Teacher Form with a Taiwanese sample. The results revealed that this scale has satisfactory internal consistency reliability, test-retest reliability, and convergent validity. The two-factor theoretical framework of the scale is supported by empirical data, with all factor loadings exceeding 0.50. Moreover, multi-group confirmatory factor analysis showed measurement invariance across gender. The researchers initially hypothesized that gender bias would emerge in this Taiwanese sample due to Confucian cultural influences. In East Asia, avoiding eye contact (and physical contact) between different genders is considered polite, as Confucius pointed out the importance of not looking at others rudely (Olberding, 2019). Thus, the researchers expected that teachers in school settings would teach gender-related social etiquette, potentially leading to gender bias in their assessments of autism-related social skills. Previous studies have shown that teachers tend to rate boys and girls differently—giving girls higher scores in general social skills and boys higher scores in general problem behaviors (Muntoni & Retelsdorf, 2018; Sasikala & Swarnakumari, 2019). However, contrary to the expectations, this study found measurement invariance across gender, suggesting that teachers’ observation and interpretations of autism-related social skill performance in school settings were consistent regardless of students’ gender.

According to the social skills deficit theory proposed by Gresham and Elliott (2008), there is a competitive relationship between social skills and problem behaviors dimensions. Hence, we hypothesized that this relationship may also be observed in this Taiwanese sample. Consistent with Gresham and Elliott’s theory (2008) and previous empirical studies (Montroy et al., 2014), the current study found a significant negative correlation between social skills and problem behaviors domains related to autism. We assumed that in Taiwanese society, social skills may also represent the ability to effectively handle social situations and help individuals achieve success in social aspects (Kamper-DeMarco et al., 2020), which might potentially reduce problem behaviors. However, although this competing relationship could be found in both Asian and Western samples, we could not compare the strength of this association across cultures because there is no Western sample in this study. Future research is recommended to explore whether the strength of these associations is the same in Western and Eastern samples.

The study was a primary step in exploring whether the SSIS Autism Spectrum Scale, part of Gresham and Elliott’s SSIS-Rating Scales developed using Western samples, is applicable in Asian countries. There is no study, to the authors’ knowledge, that has examined the construct validity of the SSIS Autism Spectrum Scale through CFA. Only two studies focused on assessing the effectiveness of interventions for children with ASD with the SSIS Autism Spectrum Scale and both reported significant improvements (Hill et al., 2017; Frey et al., 2015). Another study reported the correlation coefficients for the scores of the SSIS Autism Spectrum Scale and other measures (Channell et al., 2023). This study used CFA, a more flexible and rigorous analysis technique than Exploratory Factor Analysis (EFA), allowing researchers to determine whether their data fits the hypothesized measurement model. This study added evidence for the construct validity of the SSIS Autism Spectrum Scale, which has significant implications for the practical application of this scale. Besides, the current study used a previously validated and reliable measure as a criterion measure, thereby also helping to establish convergent validity of this scale.

However, this study also had several limitations. First, the SSIS Autism Spectrum Scale is primarily designed to support social skills training in educational settings, rather than to serve as a diagnostic tool. As such, it focuses on students’ social skills performance and does not provide a comprehensive assessment of the full range of autism symptoms. This limits the scope of behavioral characteristics measured in this study. Second, to increase the response rate for items related to autism traits, the study intentionally included a higher proportion of students with special education eligibility and ASD. This approach was necessary to ensure sufficient variability for item-level analysis, particularly for traits that are infrequent in the general population. However, it also means that the sample does not fully reflect the typical distribution of students with disabilities in Taiwanese schools. Third, the scale may have captured autism-like behaviors displayed by individuals without a formal ASD diagnosis, such as repetitive behaviors or atypical motor patterns. These characteristics may be observed in siblings of children with ASD or in students with other developmental conditions, potentially affecting the specificity of item responses. Fourth, the small test-retest sample size and the fact that all participants in the test-retest group were recruited from a single school raise concerns about the generalizability of the test-retest reliability findings. While selecting participants from a single site facilitated the follow-up process, it may also have introduced school-specific bias in the stability estimates. Fifth, to ensure adequate representation of students with special education needs or Autism Spectrum Disorder (ASD), teachers were permitted to rate multiple students. Despite ASD being one of the prevalent diagnoses in schools now, the teacher–student ratios for special educators in Taiwan remain high, with many teachers having to manage multiple students as case managers. As a result, some teachers in this study provided multiple ratings, which could introduce potential dependence that violated the assumption of independent observations. Although the number of repeated ratings per teacher was not excessive, this could affect standard errors and parameter estimates (Muthén, 1994), suggesting that results should be interpreted with more caution. However, prior research suggested that confirmatory factor analysis (CFA) is generally robust to minor violations of the independence assumption, especially when intraclass correlations are low and the sample size is adequate (Meade et al., 2008; Stapleton, 2006).

Given that this study utilized a cross-sectional research design, we suggest that future researchers explore the applicability of using this scale to track the long-term impacts of social skills training on reducing autism symptoms, using a longitudinal research design. Such investigation may lead to a better understanding of the complexities of their developmental processes and assist in the development of more appropriate social skills interventions, ultimately leading to better outcomes and well-being for children with autistic traits. Besides, it is recommended that future researchers explore the social learning needs of different autism subgroups in various settings, using this scale to record the strengths and weaknesses of these children. Autism is a highly heterogeneous condition, and children from different autism subgroups may exhibit very different strengths and face different challenges (Rao et al., 2008). Moreover, social skills are not fixed or universal, but are dynamic and context-dependent (Friedberg, 2012). Thus, this suggests that the specific social skills and behaviors that are valued and prioritized may vary in different social situations. Capturing the unique needs of different subgroups can aid in creating more specific and effective interventions in the future.

Acknowledgements

This research was funded by the Buddhist Tzu Chi Charity Foundation (Project Number: TCCCF-P-105005), Hualien County, Taiwan.

Funding

Open access funding provided by National Taiwan Normal University

Declarations

Competing interests

The authors have no financial or proprietary interests in any material discussed in this article.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Afsharnejad, B., Falkmer, M., Black, M. H., Alach, T., Lenhard, F., Fridell, A., Coco, C., Milne, K., Bölte, S., & Girdler, S. (2021). KONTAKT® social skills group training for Australian adolescents with autism spectrum disorder: A randomized controlled trial. European Child & Adolescent Psychiatry, 31, 1695–1713. 10.1007/s00787-021-01814-6 [DOI] [PubMed] [Google Scholar]
  2. Beggiato, A., Peyre, H., Maruani, A., Scheid, I., Rastam, M., Amsellem, F., Gillberg, C. I., Leboyer, M., Bourgeron, T., Gillberg, C., & Delorme, R. (2017). Gender differences in autism spectrum disorders: Divergence among specific core symptoms. Autism Research, 10(4), 680–689. 10.1002/aur.1715 [DOI] [PubMed] [Google Scholar]
  3. Chang, C., & Wu, Y. (2006). Early development and current condition in children with asperger syndrome and high-functioning autism: A comparative study. Bulletin of Special Education, 31, 139–164. [Mandarin Chinese]http://140.122.63.105/upload/journal/prog/WQO_18G0_24OY_4911022.pdf [Google Scholar]
  4. Channell, M. M., Mattie, L. J., Schworer, E. K., & Fidler, D. J. (2023). Using the social skills improvement system (SSiS) rating scales to assess social skills in youth with down syndrome. Frontiers in Psychology, 14, 1105520. 10.3389/fpsyg.2023.1105520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464–504. [Google Scholar]
  6. Cheung, P. P., Siu, A. M., & Brown, T. (2017). Measuring social skills of children and adolescents in a Chinese population: Preliminary evidence on the reliability and validity of the translated Chinese version of the social skills improvement System-Rating scales (SSIS-RS-C). Research in Developmental Disabilities, 60, 187–197. [DOI] [PubMed] [Google Scholar]
  7. Conroy, M. A., Boyd, B. A., Asmus, J. M., & Madera, D. (2007). A functional approach for ameliorating social skills deficits in young children with autism spectrum disorders. Infants & Young Children, 20(3), 242–254. 10.1097/01.IYC.0000277755.93313.7d [DOI] [Google Scholar]
  8. Crosby, J. W. (2011). Test review: FM Gresham & SN Elliott social skills improvement system rating scales. Minneapolis, MN: NCS Pearson, 2008. Journal of Psychoeducational Assessment, 29(3), 292–296. [Google Scholar]
  9. Davidovitch, M., Slobodin, O., Weisskopf, M. G., & Rotem, R. S. (2020). Age-specific time trends in incidence rates of autism spectrum disorder following adaptation of DSM‐5 and other ASD‐related regulatory changes in Israel. Autism Research, 13(11), 1893–1901. 10.1002/aur.2420 [DOI] [PubMed] [Google Scholar]
  10. Finch, H. (2014). Measurement invariance. In A. C. Michalos (Ed.), Encyclopedia of quality of life and Well-Being research. Springer. 10.1007/978-94-007-0753-5_1759 [DOI]
  11. Fombonne, E. (2018). The rising prevalence of autism. Journal of Child Psychology and Psychiatry, 59(7), 717–720. 10.1111/jcpp.12941 [DOI] [PubMed] [Google Scholar]
  12. Frey, A. J., Small, J. W., Feil, E. G., Seeley, J. R., Walker, H. M., & Forness, S. (2015). First step to success: Applications to preschoolers at risk of developing autism spectrum disorders. Education and Training in Autism and Developmental Disabilities, 50(4), 397. [PMC free article] [PubMed] [Google Scholar]
  13. Friedberg, R. D., Gorman, A. A., Wilt, L. H., Biuckians, A., & Murray, M. (2012). Cognitive behavioral therapy for the busy child psychiatrist and other mental health professionals: Rubrics and rudiments. Routledge.
  14. Gresham, F. M., & Elliott, S. N. (1989). Social skills deficits as a primary learning disability. Journal of Learning Disabilities, 22(2), 120–124. 10.1177/002221948902200207 [DOI] [PubMed] [Google Scholar]
  15. Gresham, F. M., & Elliott, S. N. (2008). Social skills improvement system (SSIS) rating scales. Pearson.
  16. Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2018). Multivariate data analysis (8th ed.). Prentice-Hall.
  17. Head, A. M., McGillivray, J. A., & Stokes, M. A. (2014). Gender differences in emotionality and sociability in children with autism spectrum disorders. Molecular Autism, 5(1), 19. 10.1186/2040-2392-5-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hill, T. L., Gray, S. A., Baker, C. N., Boggs, K., Carey, E., Johnson, C., Kamps, J. L., & Varela, E., R (2017). A pilot study examining the effectiveness of the PEERS program on social skills and anxiety in adolescents with autism spectrum disorder. Journal of Developmental and Physical Disabilities, 29, 797–808. 10.1007/s10882-017-9557-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hoang, V. M., le, T. V., Chu, T. T. Q., le, B. N., Duong, M. D., Thanh, N. M., Pham, T., Minas, V., H., & Bui, T. T. H. (2019). Prevalence of autism spectrum disorders and their relation to selected socio-demographic factors among children aged 18–30 months in Northern Vietnam, 2017. International Journal of Mental Health Systems, 13, 29–29. 10.1186/s13033-019-0285-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hsieh, M., & Wang, H. (2025). Chinese-version social skills improvement system rating scales. Chinese Behavioral Science Corporation.
  21. Hsu, C. F. (2007). A cross-cultural comparison of communication orientations between Americans and Taiwanese. Communication Quarterly, 55(3), 359–374. 10.1080/01463370701497831 [DOI] [Google Scholar]
  22. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  23. Huang, C., & Wu, Y. (2013). Autism behavior Checklist-Taiwan version. Psychological Publishing Co., Ltd.
  24. Hyman, S. L., Levy, S. E., Myers, S. M., Kuo, D. Z., Apkon, S., Davidson, L. F., Ellerbeck, K. A., Froster, J. E. A., Noritz, G. H., Leppert, M. O., Saunders, B. S., Stille, C., Yin, L., Weitzman, C. C., Childers, D. O. Jr., Levine, J. M., Peralta-Carcelen, A. D., Poon, J. K., Smith, P. J., &.Bridgemohan, C. (2020). Identification, evaluation, and management of children with autism spectrum disorder. Pediatrics, 145(1), e20193447. 10.1542/peds.2019-3447 [DOI] [PubMed] [Google Scholar]
  25. Kamper-DeMarco, K. E., Shankman, J., Fearey, E., Lawrence, H. R., & Schwartz-Mette, R. A. (2020). Linking social skills and adjustment. Social skills across the life span (pp. 47–66). Academic.
  26. Kline, R. B. (2023). Principles and practice of structural equation modeling (4th ed.). Guilford.
  27. Krug, D. A., Arick, J., & Almond, P. (1980). Behavior checklist for identifying severely handicapped individuals with high levels of autistic behavior. Child Psychology & Psychiatry & Allied Disciplines, 21(3), 221–229. 10.1111/j.1469-7610.1980.tb01797.x [DOI] [PubMed] [Google Scholar]
  28. Lai, M., Lombardo, M. V., Auyeung, B., Chakrabarti, B., & Baron-Cohen, S. (2015). Sex/gender differences and autism: Setting the scene for future research. Journal of the American Academy of Child and Adolescent Psychiatry, 54, 11–24. 10.1016/j.jaac.2014.10.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Marro, B. M., Kang, E., Hauschild, K. M., Normansell, K. M., Abu-Ramadan, T. M., & Lerner, M. D. (2019). Social performance–based interventions promote gains in social knowledge in the absence of explicit training for youth with autism spectrum disorder. Bulletin of the Menninger Clinic, 83(3), 301–325. 10.1521/bumc.2019.83.3.301 [DOI] [PubMed] [Google Scholar]
  30. Meade, A. W., Johnson, E. C., & Braddy, P. W. (2008). Power and sensitivity of alternative fit indices in tests of measurement invariance. Journal of Applied Psychology, 93(3), 568–579. 10.1037/0021-9010.93.3.568 [DOI] [PubMed] [Google Scholar]
  31. Mohan, R. (2023). Measurement, evaluation and assessment in education. PHI Learning Pvt. Ltd.
  32. Montroy, J. J., Bowles, R. P., Skibbe, L. E., & Foster, T. D. (2014). Social skills and problem behaviors as mediators of the relationship between behavioral self-regulation and academic achievement. Early Childhood Research Quarterly, 29(3), 298–309. 10.1016/j.ecresq.2014.03.002 [DOI] [Google Scholar]
  33. Muntoni, F., & Retelsdorf, J. (2018). Gender-specific teacher expectations in reading—the role of teachers’ gender stereotypes. Contemporary Educational Psychology, 54, 212–220. 10.1016/j.cedpsych.2018.06.012 [DOI] [Google Scholar]
  34. Muthén, B. O. (1994). Multilevel covariance structure analysis. Sociological Methods & Research, 22(3), 376–398. 10.1177/0049124194022003006 [DOI] [Google Scholar]
  35. Nader, D., Dunlop, B., Tomaszewski, B., Sam, A. M., Hume, K., Aljuaid, M., & Odom, S. L. (2023). Exploratory factor analysis of the social skills improvement system (SSIS) for autistic children: Implications for use and educational placement. Research in Autism Spectrum Disorders, 104, 102144. [Google Scholar]
  36. Nangle, D., Grover, R. L., Holleb, L. J., Cassano, M., & Fales, J. (2010). Defining competence and identifying target skills. In D. W. Nangle D. J. Hansen, C. A. Erdley, & P. J. Norton (Eds.), Practicioner’s guide to empirically based measures of social skills. (pp. 3–19). Springer. https://www.researchgate.net/publication/251224552_Defining_Competence_and_Identifying_Target_Skills
  37. Olberding, A. (2019). The wrong of rudeness: Learning modern civility from ancient Chinese philosophy. Oxford University Press.
  38. Olmeda, R. E., & Kauffman, J. M. (2003). Sociocultural considerations in social skills training research with African American students with emotional or behavioral disorders. Journal of Developmental and Physical Disabilities, 15, 101–121. 10.1023/A:1022871232435 [DOI] [Google Scholar]
  39. Panayiotou, M., Santos, J., Black, L., & Humphrey, N. (2022). Exploring the dimensionality of the social skills improvement system using exploratory graph analysis and bifactor-(S– 1) modeling. Assessment, 29(2), 257–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance conventions and reporting: The state of the Art and future directions for psychological research. Developmental Review, 41, 71–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rao, P. A., Beidel, D. C., & Murray, M. J. (2008). Social skills interventions for children with Asperger’s syndrome or high-functioning autism: A review and recommendations. Journal of Autism and Developmental Disorders, 38, 353–361. 10.1007/s10803-007-0402-4 [DOI] [PubMed] [Google Scholar]
  42. Sasikala, G., & Swarnakumari, D. P. (2019). Parent and teacher ratings on gender differences in social skills, problem behaviours and academic competence of children with mild intellectual disability in inclusive education. International Journal of Research and Analytical Reviews, 6(2), 599–602. 10.1016/j.sbspro.2010.07.256 [DOI] [Google Scholar]
  43. Sedgewick, F., Hill, V., Yates, R., Pickering, L., & Pellicano, E. (2016). Gender differences in the social motivation and friendship experiences of autistic and non-autistic adolescents. Journal of Autism and Developmental Disorders, 46, 1297–1306. 10.1007/s10803-015-2669-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stapleton, L. M. (2006). An assessment of practical solutions for structural equation modeling with complex sample data. Structural Equation Modeling, 13(1), 28–58. 10.1207/s15328007sem1301_2 [DOI] [Google Scholar]
  45. Supekar, K., & Menon, V. (2015). Sex differences in structural organization of motor systems and their dissociable links with repetitive/restricted behaviors in children with autism. Molecular Autism, 6, 50. 10.1186/s13229-015-0042-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Taber, K. S. (2018). The use of Cronbach’s alpha when developing and reporting research instruments in science education. Research in Science Education, 48, 1273–1296. 10.1007/s11165-016-9602-2 [DOI] [Google Scholar]
  47. Wang, H. T., Sandall, S. R., Davis, C. A., & Thomas, C. J. (2011). Social skills assessment in young children with autism: A comparison evaluation of the SSRS and PKBS. Journal of Autism and Developmental Disorders, 41, 1487–1495. [DOI] [PubMed] [Google Scholar]
  48. Wood-Downie, H., Wong, B., Kovshoff, H., Cortese, S., & Hadwin, J. A. (2021). Research review: A systematic review and meta‐analysis of sex/gender differences in social interaction and communication in autistic and nonautistic children and adolescents. Journal of Child Psychology and Psychiatry, 62(8), 922–936. [DOI] [PubMed] [Google Scholar]
  49. Zeidan, J., Fombonne, E., Scorah, J., Ibrahim, A., Durkin, M. S., Saxena, S., Yusuf, A., Shih, A., & Elsabbagh, M. (2022). Global prevalence of autism: A systematic review update. Autism Research, 15(5), 778–790. 10.1002/aur.2696 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Autism and Developmental Disorders are provided here courtesy of Springer

RESOURCES