Assessing Social Skills in Early Elementary-Aged Children With Autism Spectrum Disorders

Abstract

This study employed a newly developed measure, the Social Skills Q-Sort (SSQ), to assess paraprofessionals’ and teachers’ reports of social skills for children with and without ASD. Paraprofessionals and teachers showed good rater-agreement on the SSQ. ROC curve analyses yielded an excellent profile of sensitivity and specificity for discriminating between children with ASD and typically developing children. The paraprofessional SSQ converged with objective ratings of playground social behavior; however, there was little evidence of convergence between SSQ scores and parent and teacher ratings on questionnaire measures. The SSQ may be effective in screening for ASD and the severity of ASD-related social communication challenges.

Keywords

autism social skills assessment social involvement social networks

Diagnosing and classifying children with autism spectrum disorder (ASD) has been refined using the gold-standard Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000) and the Autism Diagnositic Interview-Revised (ADI-R; Lord, Rutter, & Le Couteur, 1994). Although it is expected that most children with autism will be identified before they enter school, many children go undiagnosed particularly if they are performing academically (Brock, Jimerson, & Hansen, 2006). While there are many avenues in which families can turn to receive a diagnosis, many families (especially those in urban settings) rely heavily on school resources; however, many schools are ill-equipped to provide screening or diagnostic services for children suspected of having an ASD, in part, because these assessments require an expertly trained clinician or psychologist to administer, which can be time-consuming, quite costly, and unrealistic for public school settings. Schools need a more cost-efficient and easily administered way of assessing and screening children’s social skills, a core deficit of ASD. While existing paper and pencil rating systems for social functioning reliably measure specific domains of social functioning, they do not fully capture the gamut of social skills that children with ASD display. Because autism is a spectrum disorder, children’s social abilities are difficult to measure and multiple domains of social functioning need to be evaluated. Thus, this study introduces a new measure, the Social Skills Q-Sort (SSQ), that has the potential to address these issues.

The ability to understand social cues and engage in interpersonal exchanges requires a set of social skills, which is inherent in most typically developing children. In contrast, children with ASD experience challenges in social reciprocity and communication that may impede their ability to navigate social interactions, which in turn, may interfere with their ability to establish lasting and quality friendships. Children with ASD often experience challenges surrounding social pragmatics (e.g., turn-taking in conversation, initiating conversation, and the ability to take the listener’s perspective), perseverative speech, and emotion regulation, expression, and understanding (Williams White, Keonig, & Scahill, 2007). These deficits may possibly lead to rejection and isolation from peers (Bellini, 2004), particularly when integrated with typically developing children in inclusive school settings (Chamberlain, Kasari, & Rotherham-Fuller, 2007; Kasari, Locke, Gulsrud, & Rotheram-Fuller, 2011). Inclusion of children with ASD in mainstreamed classrooms has become an increasingly common practice (Kasari, Freeman, Bauminger, & Alkin, 1999); however, inclusion alone does not improve social functioning or peer relationships (Kasari et al., 2011). In fact, children with ASD are often on the periphery of their classroom social networks (Chamberlain et al., 2007; Kasari et al., 2011) and their social involvement within these networks typically worsens as they progress through elementary school (Rotheram-Fuller, Kasari, Chamberlain, & Locke, 2010).

Although researchers have begun to describe the developmental course of social interactions and relationships among children with ASD (McConnell, 2002), measuring social functioning is difficult particularly in higher functioning children with ASD. Although children with ASD included in general education classrooms have comparable cognitive ability scores within the average to above average range and are performing “at grade level” with respect to their typically developing classmates, there is a disconnect in their social competencies (Kasari et al., 2011; Rotheram-Fuller et al., 2010). Since social skills encompass several domains of social functioning (e.g. communication, friendship, engagement, play, emotion recognition/understanding, etc.) that vary across children with ASD, reliably and comprehensively measuring children’s social skills sets can be difficult to achieve when using only one approach, methodology, or rater (Campbell & Fiske, 1959).

Aside from the ADOS and the ADI, the gold-standard instruments used to classify children with ASD, at present, there are only a few paper and pencil assessment tools that are used to identify social skills precisely and reliably (Bellini & Hopf, 2007). Among the most frequently used for children with ASD are the Social Skills Rating System (SSRS; Gresham & Elliott, 1990), the Social Responsiveness Scale (SRS; Constantino & Gruber, 2005) and the Autism Social Skills Profile (ASSP; Bellini & Hopf, 2007). Although several studies have used the SSRS to measure social skills in children with autism (Macintosh & Dissanayake, 2006; Meier, DiPerna, & Oster, 2006), the SSRS was not specifically designed to assess children with autism. Thus, few of the items address the unique patterns of social behavior exhibited by children with autism (Bellini & Hopf, 2007), providing an insufficient profile of these children’s social skills and behaviors. Another frequently used social skills assessment is the SRS (Constantino & Gruber, 2005). While the SRS yields a single total score with higher values indicating greater severity of autism spectrum symptoms, the existence of separate subdomains of social behavior have not been empirically supported (Bellini & Hopf, 2007; Pine, Luby, Abbacchi, & Constantino, 2006). Thus, it may be difficult to parse out specific areas of social deficits in children with ASD. More recently, Bellini & Hopf (2007) developed the ASSP, a questionnaire specifically designed for children and adolescents with autism that provides a comprehensive measure of social functioning. While the ASSP provides a description of children’s social functioning, it was developed as an intervention planning tool and has not been used a school-based screening tool.

Paper and pencil measures have traditionally been given to parents and teachers to complete who may or may not be privy to children’s relationships with peers at school or have the opportunity to observe children in the context in which they apply these skills (i.e. the playground). Since paraprofessionals spend a considerable amount of time with children with ASD during social periods such as recess on the playground, they may be the most informed raters of children’s social abilities. In addition, raters often do not have the ability to distinguish frequency from significance using these instruments, thereby limiting the specificity and accuracy of identifying areas of strength and weakness in children’s skill sets (Waters & Deane, 1985). Thus, alternative measurement strategies such as the Q-sort methodology may elucidate specific domains of children’s social abilities to provide a more thorough profile of children’s overall social competence.

The Q-sort methodology allows raters to rank order, from most characteristic to most uncharacteristic, a set of descriptive items in which scores are to be assigned (Waters & Deane, 1985). This methodology (described below) offers several advantages to paper and pencil checklists. First, raters are kept unaware of the constructs that will be scored from the data they provide, thereby reducing response bias. Second, raters are not required to have detailed knowledge of norms for each item since they are describing one specific child. Third, the significance of a behavior is clearly distinguished from the frequency with which it occurs since raters sort items into a rank order in the context of a well defined set of other items (Waters & Deane, 1985).

In the development of any instrument, a gold-standard measure should be used to cross-validate the new rating scale. Although time intensive and costly, behavioral observations are well regarded in capturing children’s social abilities in applied situations such as the classroom and playground (Kasari et al., 2011) and serve as a gold standard for assessing severity of social communication problems in daily life. The Q-sort developed in this study may be more cost-effective and efficient in providing detailed information on children’s social abilities and friendships and does not require a trained clinician, psychologist, or researcher to administer as with the ADOS. To ascertain its convergent validity, Q-Sort ratings were cross-validated against detailed playground behavioral observations, which served as the “gold standard” in this study.

The objective of this study was to explore various aspects of children’s social functioning as rated by paraprofessionals and teachers. This study employed the SSQ to examine the social skills of elementary-aged children with ASD and a group of typically developing matched peers. We hypothesized that the SSQ would show evidence of rater agreement, and convergent validity with behavioral observations. We also explored convergence with checklist measures of children’s social skills. Furthermore, paraprofessionals’ and teachers’ ratings from the SSQ for children with ASD were expected to be significantly lower than their ratings of the matched peers.

Method

Participants

There were 42 participants in this study: 21 children with ASD and 21 matched typically developing children and their parents. Children were included for the ASD group if they were referred from their school administrations and had a diagnosis of ASD from a licensed psychologist, were fully included in a regular education classroom for at least 80% of the school day, were between the ages of 5 to 8 years old, and were in kindergarten through second grades.

Children in the matched group were recruited from the same classrooms as children with ASD. Twenty-one children were matched on age, gender (predominantly male; 86%), and grade to serve as a comparison group. Children were not matched on mental age because intelligence tests were prohibited in the participating schools.

Table 1 summarizes the background characteristics for children with ASD and their matched peers. In addition, 17 paraprofessionals (88% female) and 19 teachers (100% female) participated in this study. Children in classrooms with more than one child with ASD shared the same paraprofessional. Both paraprofessionals and teachers were from diverse ethnic backgrounds (Paraprofessionals: 23.5% Caucasian, 17.6% African American, 52.9% Latino, and 5.9% Asian; Teachers: 36.8% Caucasian, 15.8% African American, 21.1% Latino, 15.8% Asian, and 10.5%, Other). Overall, teachers had an average of 12.26 ± 8.28 years of teaching experience and the majority (63.2%) had some training (i.e. workshops, consultations, seminars, graduate courses, and professional development) in working with children with ASD. A smaller percentage of teachers (31.6%) were formally trained in special education. Further demographic information was not collected on the 17 paraprofessionals that participated in this study.

Table 1.

Child Demographics.

	ASD (N = 21) M (SD) / Frequency (%)	Matched peers (N = 21) M (SD) / Frequency (%)
Gender
Male	18 (85.7)	18 (85.7)
Female	3 (14.3)	3 (14.3)
Grade
Kindergarten	4 (19)	4 (19)
First	11 (52.4)	11 (52.4)
Second	6 (28.6)	6 (28.6)
Ethnicity
Caucasian	9 (42.9)	7 (33.3)
Minority	12 (58.1)	14 (66.7)
Chronological age (years)	6.71 (1.00)	6.67 (0.86)

Measures

Social Skills Q-Sort (Locke, Kretzmann, & Kasari, 2008)

Consistent with Waters and Deane (1985), the SSQ was developed and revised in four stages. In the first stage, the social skills and autism-related intervention literature was reviewed. From this literature review, a preliminary item set was constituted. The items were selected to represent the gamut of social skills since many school-aged children with ASD experience a range of enduring and complex deficits that fall underneath the larger umbrella of social functioning which encompasses a broad range of abilities including foundational skills (e.g. using eye contact, playing with an object, taking turns with peers) as well as more complex skills (e.g. having reciprocal conversations, playing social-interactional games with typically developing peers, establishing, and maintaining friendships). In the second stage, the preliminary item set was distributed to professionals (i.e. clinicians, teachers, and researchers) working with children with ASD. They were asked to rate a child with ASD that they knew well. These professionals were also instructed to document any items that were unclear or irrelevant. In the third stage, behaviors that never occurred, could not be sorted with good agreement, or had very little variance across subjects were revised or eliminated. In the final stage, 43 professionals (i.e. clinicians, teachers, and researchers) were given the SSQ and asked to sort the items (on cards) on a hypothetical typically developing child in elementary school who embodies the “ideal socially competent child.” Each expert rater was independently given the SSQ to complete at a time that was convenient to them. Their average Q-sort ratings of this hypothetical child were used as a standard criterion against which Q-sort ratings obtained in this study were compared. Simple correlations were conducted between the SSQ criterion means for each item from the expert sort and each item from the paraprofessional and teacher reported sorts for both children with ASD and their matched peers. The correlation coefficient between the ideal construct definition and the description of the subject was used as the subject’s SSQ score. The correlation coefficients ranged from −1.00 to & +;1.00, where positive correlations indicated closer approximations to the criterion—the ideally socially competent elementary-aged child – and negative correlations indicated deviance from the criterion (Waters & Deane, 1985).

The SSQ was designed to permit descriptions of individual children’s repertoire of social skills in order to create one comprehensive social skills profile. The SSQ consists of a set of 100 descriptive cards to be arranged or “sorted” into nine piles whose designations range from most characteristic (pile 9) to least characteristic (pile 1) of a particular subject. Raters sorted the items into three piles and then subdivided these piles into a total of nine. Then, working from the outer piles toward the center, each pile was adjusted so that the final sort conforms to a symmetrical, unimodal distribution with specified numbers of items in each of the nine piles (e.g. 5, 8, 12, 16, 18, 16, 12, 8, 5). Items may not be deleted or excluded from the sort. Extremely placed items were the most salient and most informative in conveying understanding of the child being described. Items placed in the middle categories can be of three kinds: (a) the particular item is of little salience for understanding the child; (b) the particular item is sometimes true and sometimes not true—expressing its average importance; and (c) the rater has little or no information or is uncertain about the behaviors or qualities involved. Each item on the SSQ was scored in terms of its placement (piles 1-9) in the distribution (e.g., each of the five items in pile 9 received a score of 9; each of the eight items in pile 8 received a score of 8, and so forth). Subsequently, the items were correlated with the criterion means to generate one total SSQ score. The specific items on the SSQ as well as their criterion means can be found in Appendix A.

Playground Observation of Peer Engagement (POPE) (Kasari, Rotheram-Fuller, & Locke, 2005; Kasari et al., 2011)

The POPE is a timed-interval behavior coding system that examines children’s engagement and social communicative behaviors with peers on the playground. Independent observers watched the child with ASD on the playground for 40 consecutive seconds and then coded for 20 s for 10 min during the recess or lunch period on two separate occasions within 1 week. Observers were trained and considered reliable with a criterion α > 0.80. Playground engagement states were summed and divided by the total number of observed intervals. Reliability was estimated with Kappa statistic and averaged to be .91 (range .83-.96; see Kasari et al., 2011).

Social Skills Rating System—Teacher and Parent Forms (SSRS; Gresham & Elliott, 1990)

The SSRS—teacher form is a 57-item assessment of children’s social behaviors. The first 48 items are measured on a 0 to 2 point scale ranging from “never” to “very often”, whereas the remaining nine items are rated on a scale from 1 to 5 ranging from “lowest 10%” to “highest 10%”. The teacher-rated SSRS yields three categories: Social Skills, Problem Behaviors, and Academic Competence. The SSRS—parent form is a 55-item assessment of children’s social skills and problem behaviors. The same 0 to 2 point scale is used. The first 38 items yield the Social Skills domain, and items 39 to 55 yield the Problem Behaviors domain. The SSRS has high internal consistency (α = .73-.94) and test-retest reliability (r > .65-.93; Gresham & Elliott, 1990).

Social Responsiveness Scale (SRS; Constantino & Gruber, 2005)

The SRS is a 56-item rating scale that assesses the level of reciprocal social behavior characteristically demonstrated by children with ASD. The SRS has been widely used to distinguish children with a pervasive developmental disorder from those with other psychiatric conditions (Constantino et al., 2006). The measure is used for children between the ages of 4 to 18 years with internal consistency of .88, inter-rater reliability of .75, and test–retest reliability of .88 (Constantino & Gruber, 2005).

Procedures

Participants were recruited from 19 classrooms in six elementary schools in an urban area. Once a letter of collaboration was obtained from the principal, recruitment materials were provided to the school administration to distribute to interested families. Upon return, a peer consent and parent packet were sent home to a randomly selected matched peer. During this visit, an independent observer collected a behavioral observation on the child with ASD on the playground during recess or lunch.

Paraprofessionals and teachers were asked to rate the child with autism and his/her matched peer on the SSQ in order to test inter-rater agreement. Administration of the SSQ was counterbalanced where half of the paraprofessionals first gave Q-sort ratings to the child with autism and the other half first gave Q-sort ratings to the matched peer. Teachers were also asked to complete the SSRS and SRS for both the children with autism and their matched peers. As an incentive for their time, paraprofessionals and teachers were offered a small stipend.

Results

Paraprofessionals’ and Teachers’ Q-Sorts Versus Criterion

The degree of association between the ideal criterion ratings and the SSQ ratings of each child in this study was calculated for all paraprofessional- and teacher-reported SSQs. The mean for the paraprofessional-reported SSQs for children with ASD was 0.15 ± 0.41 with a range of −.54 to 0.70 and 0.62 ± 0.19 with a range of 0.02 to 0.80 for the matched peers. Similarly, the mean for the teacher-reported SSQs for children with ASD was 0.07 ± 0.36 with a range of −0.49 to 0.62 and 0.62 ± 0.29 with a range of −0.44 to 0.84 for the matched peers. A rater (i.e. paraprofessionals and teachers) by group (i.e. autism and matched peers) repeated measures ANOVA (RMANOVA) indicated there were no significant within group differences between paraprofessional and teacher ratings on the SSQ, F (1, 38) = .75, p = .39. See Figures 1 and 2. However, the RMANOVA yielded significant between-group effects in paraprofessional and teacher ratings on the SSQ, F (1, 38) = 30.72, p < .0001.

Figure 1.

Distribution of paraprofessional and teacher-rated SSQs for children with ASD.

Figure 2.

Distribution of paraprofessional and teacher-rated SSQs for the matched sample.

Receiver Operating Characteristic (ROC) curve analysis indicated that the optimal cutpoints for identifying children with ASD were −.19 on the teacher SSQ (sensitivity, .71; specificity, .95) and −.17 on the paraprofessional SSQ (sensitivity, .75; specificity, 1.00). Hence, at these cutpoints, the SSQ was able to detect the majority of true positive cases of ASD and made almost no errors in falsely identifying a typical child as having ASD.

Rater Agreement

Intra-class correlations (ICC) were conducted to examine rater agreement between paraprofessionals’ and teachers’ reports of social functioning on the SSQ for children with ASD and their matched peers. Overall inter-rater agreement between paraprofessionals’ and teachers’ ratings of social skills for children with ASD was good, ICC = .74. Similar agreement was attained for the matched peers, ICC = .71.

Convergent and Discriminant Validity

Correlations between paraprofessional-reported SSQ scores and independent observer-rated playground engagement scores for children with ASD are reported in Table 2. There was some evidence of convergent validity for paraprofessional-rated SSQs. Correlations were significant between paraprofessional SSQ scores and (a) observer-rated joint engagement and games, (b) children’s total initiations toward peers, and (c) children’s appropriate responses to peers’ initiations. However, paraprofessional-rated SSQs were not significantly correlated with children’s solitary engagement, successful initiations and total responses to peers (range r = −.46-.29). In addition, teacher rated SSQs had negligible correlations with observer rated playground engagement and social communicative behaviors (range r = −.12-.40; not presented in Table 2).

Table 2.

Correlations Between SSQ, Observational Proportion Scores, and SRS and SSRS Scores.

Converging factors	Paraprofessional SSQ ASD	Teacher SSQ ASD
Observer-rated
Joint-engagement & games	.48*	.27
Initiations	.51*	.40
Appropriate responses	.60*	.17
Solitary engagement	−.46	−.12
Successful initiations	.29	.07
Responses	.22	−.05
Parent-rated
SSRS social skills	.30	.04
SRS total score	−.15	.00
Teacher-rated
SSRS social skills	.42	.64**
SRS total score	−.21	−.64**
Diverging factor
Teacher-rated
SSRS academic competence	.23	.24

Note: *p<.05. **p<.01.

Correlations between parents’ and teachers’ ratings were not significant for both the SRS (r_ASD = .05. p = .82 and r_MATCHED = −.18, p = .44) and the SSRS (r_ASD = −.02, p = .92 and r_MATCHED = −.18, p = .44). We also examined paraprofessional- and teacher-reported SSQs and parent- and teacher-rated SSRS and SRS scores (see Table 2). Overall, SSQ scores did not evidence a statistically significant association with parent- and teacher-rated SRS and SSRS scores for children with ASD.

Predicting to Playground Observational Ratings

In contrast to teachers’ ratings on the SSRS and SRS, parents’ questionnaire ratings yielded significant correlations with one variable from the POPE, appropriate responses, r_SSRS = .54 and r_SRS = −.50, respectively. A stepwise multiple regression was conducted to determine if paraprofessional-rated SSQs and parent ratings on the SSRS and SRS contributed independent variance to children’s appropriate responses on the POPE. The results indicated that paraprofessional-rated SSQs significantly predicted children’s appropriate responses to peers on the playground, ß = .60, t(15) = 2.87, p = .01. Parents’ ratings on the SSRS and SRS did not contribute additional variance to children’s appropriate responses to peers on the playground above and beyond paraprofessional-rated SSQs.

Discussion

This study had two overarching goals. First, the SSQ was developed and paraprofessionals’ and teachers’ ratings of children’s social skills were compared for agreement on young children with and without ASD as well as against criterion ratings. Second, the validity of the SSQ was evaluated. The results demonstrated the potential utility of the Q-sort methodology for deriving a broad school profile of children’s social skills. Both teacher- and paraprofessional-rated SSQ profiles were sensitive to the presence of ASD, with good sensitivity and excellent specificity. The findings also showed that paraprofessional-rated SSQs were associated with objective observations of children’s social behavior during free play at recess, offering critical evidence of convergent validity. However, teacher-rated SSQs showed little evidence of convergent validity. Hence, the discussion focuses on the capacity of the paraprofessional-rated SSQ to assess social skills in children with ASD, and also addresses possible reasons for the lack of evidence of validity for the teacher-reported SSQ.

These results provide further evidence of the disparity in social functioning between children with ASD and their typically developing peers from a unique and often neglected perspective—that of paraprofessionals and teachers. Although social skills checklists are economical and efficient, they may not fully capture multiple areas of children’s social functioning in applied contexts. The SSQ may fill this void and act as a reliable and cost-effective school-based screening tool in identifying children who may need evaluation for an ASD.

There was evidence suggesting that paraprofessional-rated SSQ scores converged with objective observations of children’s social engagement and behavior on the playground. It is evident that generalization of skills to other contexts is challenging for children with ASD (McConnell, 2002), and the presence (or lack of) social skills may be more apparent in unstructured settings such as the playground (Kasari et al., 2011). The fact that paraprofessionals’ ratings on the SSQ converged with observer ratings on the POPE more than teachers’ and parents’ ratings lend credence to the idea that paraprofessionals are more accurately assessing children’s social skills in the playground context. However, paraprofessionals and teachers had high inter-rater reliability for both samples traversing the classroom and the playground indicating that the social skills being measured by the SSQ may be situation specific to the school context.

Although paraprofessionals and teachers had high agreement in rating children’s social skills on the SSQ, there was little evidence of convergent validity for the teacher-rated SSQ, which may be due to several factors. First, these findings may be limited because of the small sample size in each group, thereby limiting statistical power to detect a relationship. Second, the observed correlations between teacher-rated SSQs and observers’ ratings of children’s playground engagement may be due to the fact that teachers often do not spend time with children on the playground during recess and lunch. These periods are typically monitored by paraprofessionals, which may explain the lack of convergence between teachers and observers.

In considering why neither paraprofessional- nor teacher-rated SSQ scores converged with parent-rated checklist scores, it is noted that items on the SSQ may not capture the same social skills as the SSRS and SRS. The SSQ was created to represent the gamut of social behaviors that characterize children’s social functioning at school. In this study, both parent-reported and teacher-reported SSRS and SRS did not correlate for the sample with ASD or the matched sample, which suggests that even validated social skills instruments may be situation specific, where parents and teachers are seeing different social behaviors across the home and school contexts. While this could be a sample size issue, it is more likely that the lack of convergence between the paraprofessional- and teacher-rated SSQs and the parent- and teacher-rated SSRS and SRS indicates that children’s social skills in one domain may not generalize to other contexts. Parents and teachers may have different opinions on children’s behaviors or what behaviors they see in children resulting in low agreement in their perceptions of children’s social abilities (Reisinger et al., 2010). Thus, if parents’ and teachers’ ratings on these validated measures did not correlate in this study, it is unlikely that the SSQ would correlate with either measure.

Limitations

While this study is unique in its multimethod approach to understanding social skills in children with ASD, there were several limitations to this study. First, only 21 children with ASD and 21 matched peers were enrolled in this study. Additional research participants would increase statistical power. Second, although children with ASD recruited for this study were identified by their school districts as having an ASD and referred to this study by school personnel, research personnel did not independently confirm children’s diagnoses of ASD. Furthermore, the matched peers were not screened for ASD to rule out the possibility of sample contamination.

Furthermore, the cross-sectional design of this study and restrictions in grade level (K-2) also limit understanding of what changes might occur across age groups, classrooms, and time. Longitudinal studies may help specify a developmental trajectory in children’s social skill sets that may specify when exact difficulties arise for children with ASD that might be optimal intervention periods. Understanding how children’s social skills change as children age would help inform interventions to target developmentally appropriate areas in children’s social functioning that could potentially alleviate social difficulties at school.

Future Directions

Because the SSQ measures broad areas of children’s social functioning (see appendix), future studies should examine whether the SSQ can be used as an instrument to identify potential treatment areas for school-based social skills interventions. Researchers are only beginning to explore and truly understand the social functioning of included school-aged children with ASD and how these skills impact their relationships and engagement with peers on the playground (Rotheram-Fuller et al., 2010; Kasari et al., 2011). While a variety of school-based social skills interventions have been employed to remediate children’s social skills, the methods in which they identify children’s intervention targets are less clear (Bellini, Peters, Benner, & Hope, 2007).

In addition, future studies should examine the stability of the SSQ over time to determine whether it could be used to monitor intervention progress in children with ASD. Many studies rely on one or two responders (e.g. parents and/or the child) to provide detailed information on the social and emotional functioning of children with ASD (Bauminger & Kasari, 2000); however, the social lives of children with ASD are often complex and may not be fully understood from a single source or within a single context (Kasari et al., 2011). Hence, measuring treatment outcomes may be limited and intervention effectiveness may not be adequately captured. An essential piece to understanding children’s social skills sets is to incorporate multiple agents in the child’s life to cross-validate the information provided. The SSQ has the potential to be used by multiple raters within the school context. Future studies that explore the potential uses of the SSQ may provide a means in which interventionists plan treatment targets and measure associated outcomes.

Footnotes

Appendix

Items on the Social Skills Q-Sort	Mean
1. Prefers nonverbal methods of communication.	4.21
2. Joins group activities without being told to.	6.42
3. Is warm and responsive.	7.37
4. Gets along well with other children.	7.65
5. Is admired and sought out by other children.	6.51
6. Is helpful and cooperative.	7.02
7. Seeks physical contact with others	4.93
8. Responds to others verbal initiations.	6.49
9. Develops genuine and close relationships.	8.35
10. Avoids social situations with other children.	2.63
11. Has a stable peer group.	6.93
12. Has tantrums when doesn’t get his/her way.	2.35
13. Seeks negative attention.	3.07
14. Tries to do everything exactly right or perfect.	4.51
15. Child is rejected or not well-liked by other children.	2.40
16. Child tends to be pleased with and proud of his/her products and accomplishments.	6.44
17. Child is physically aggressive.	1.65
18. Expresses negative feeling towards peers directly and openly.	4.28
19. Child is off-topic in speech; rambles.	3.63
20. Has low self-esteem.	3.37
21. Has a reciprocal best friend.	7.21
22. Tries to manipulate others.	3.26
23. Is fearful and anxious.	3.26
24. Makes friends easily.	7.33
25. Has multiple (3 or more) reciprocal friendships with age-mates.	7.23
26. Enjoys physical activity (e.g. running, climbing, walking).	5.30
27. Is visibly deviant from peers in appearance, size, or physical condition	4.09
28. Is vital, energetic, lively.	6.21
29. Is polite and respectful of others.	6.84
30. Stares or gazes off to space.	3.72
31. Shows recognition to the feelings of other; is empathic.	7.86
32. Tends to give, lend, and share with others.	7.23
33. Cries easily.	3.56
34. Is restless and fidgety.	4.14
35. Is inhibited and constricted.	3.53
36. Accepts friends’/peers’ ideas for playing.	6.93
37. Prefers to interact with adults.	4.23
38. Has set routines.	4.74
39. Tends to become rigidly repetitive or immobilized.	2.51
40. Is curious and exploring, eager to learn, open to new experiences.	7.23
41. Perseverates in activities/topics of speech, etc.	3.02
42. Is receptive to other children.	6.81
43. Can recoup or recover after stressful experiences.	6.72
44. When in conflict or disagreement with other tends to yield and give in.	4.79
45. Tends to withdraw and disengage from others.	2.70
46. Tends to go to pieces under stress, becomes rattled, and disorganized.	2.79
47. Enjoys a multitude of activities and games	6.37
48. Is short tempered; easily enraged.	2.60
49. Engages in self-stimulatory behaviors and/or rituals	2.56
50. Easily changes from one activity to another.	6.35
51. Is agile and well-coordinated.	5.42
52. Becomes easily embarrassed.	4.40
53. Has random verbal outbursts.	2.93
54. Has rapid shifts in mood; is emotionally volatile.	2.26
55. Has peers who are friendly and welcoming.	6.26
56. Has difficulty relating to peers.	2.79
57. Is agreeable.	6.40
58. Is emotionally expressive (facially, gesturally, or verbally).	6.56
59. Sustains engagement in games.	6.63
60. Becomes anxious when the environment is unpredictable or poorly structured.	3.60
61. Has an aide, assistant, adult supervisor during social times	3.33
62. Is obedient and compliant.	5.65
63. Prefers table top games or activities.	5.02
64. Is calm and relaxed, easy-going.	6.30
65. Initiates play/games to others.	7.09
66. Child is attentive and able to concentrate.	6.53
67. Child is responsive to new ideas/suggestions.	6.91
68. Responds to direct instruction.	5.49
69. Is verbally fluent; can express ideas well in language.	6.91
70. Is awkward in turn taking interactions with peers.	3.47
71. Looks to adults for help and direction.	5.16
72. Is nice to other children.	7.09
73. Enjoys talking to other children.	6.79
74. Has negative thoughts; easily discouraged.	3.63
75. Child is cheerful.	6.42
76. Understands limits.	6.47
77. Reflective of self. Aware of his/her differences.	6.42
78. Is easily offended; sensitive to ridicule or criticism.	3.70
79. Avoids eye contact.	3.21
80. Teases other children.	2.88
81. Optimistic or positive in thinking.	6.84
82. Becomes easily motivated.	6.44
83. Tends to self-regulate effectively.	7.33
84. Child has average or above average athletic ability.	5.30
85. Child is verbally aggressive.	2.42
86. Likes to be by him/herself, enjoys solitary activities.	4.12
87. Initiates conversation to others.	6.91
88. Becomes obsessive with others.	2.40
89. Child is competent, skillful, agile, or adept.	6.30
90. Child is stubborn and inflexible.	2.98
91. Is inappropriate in emotional behavior.	2.21
92. Child appears likeable to peers and/or is pleasant to be around.	7.33
93. Yells at others.	2.98
94. Tends to be sulky or whiny.	3.63
95. Overacts to minor frustrations; is easily irritated and/or angered.	2.53
96. Understands/follows game rules.	6.28
97. Engages in solitary imaginary/fantasy play.	4.77
98. Has difficulty making social connections.	2.67
99. Engages in games with other children.	6.40
100. Is picked on by other children.	3.19

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This study was funded in part by a UCLA Dissertation Year Fellowship to the first author.

References

Bauminger

Kasari

(2000). Loneliness and friendship in high-functioning children with autism. Child Development, 71, 447-456.

Bellini

(2004). Social skill deficits and anxiety in high-functioning adolescents with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 19, 78-86.

Bellini

Hopf

(2007). The development of the autism social skills profile: A preliminary analysis of psychometric properties. Focus on Autism and Other Developmental Disabilities, 22(2), 80-87.

Bellini

Peters

J. K.

Benner

Hope

(2007). A meta-analysis of school-based social skills interventions for children with autism spectrum disorders. Remedial and Special Education, 28, 153-162.

Brock

S. E.

Jimerson

S. R.

Hansen

R. L.

(2006). Identifying, assessing, and treating autism at school. New York, NY: Springer.

Campbell

D. T.

Fiske

D. W.

(1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81-105.

Chamberlain

Kasari

Rotheram-Fuller

(2007). Involvement or isolation? The social networks of children with autism in regular classrooms. Journal of Autism and Developmental Disorders, 37, 230-242.

Constantino

J. N.

Gruber

C. P.

(2005). The social responsiveness scale (SRS) manual. Los Angeles, CA: Western Psychological Services.

Constantino

J. N.

Lajonchere

Lutz

Gray

Abbacchi

McKenna

Todd

R. D.

(2006). Autistic social impairment in the siblings of children with pervasive developmental disorders. American Journal of Psychiatry, 163, 294-296.

10.

Gresham

F. M.

Elliott

S. N.

(1990). Social skills rating system. Circle Pines, MN: American Guidance Service.

11.

Kasari

Freeman

S. F. N.

Bauminger

Alkin

M. C.

(1999). Parental perspectives on inclusion: Effects of autism and Down syndrome. Journal of Autism and Developmental Disorders, 29, 297-305.

12.

Kasari

Locke

Gulsrud

Rotheram-Fuller

(2011). Social networks and friendships at school: Comparing children with and without autism. Journal of Autism and Developmental Disorders, 41, 533-544.

13.

Kasari

Rotheram-Fuller Locke

(2005). The development of the Playground Observation of Peer Engagement (POPE) measure. Unpublished manuscript, University of California Los Angeles, Los Angeles, CA.

14.

Locke

Kretzmann

Kasari

(2008). The Social Skills Q-Sort. Unpublished manuscript, University of California Los Angeles, Los Angeles, CA.

15.

Lord

Rutter

Le Couteur

(1994). Autism Diagnostic Interview-Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism & Developmental Disorders, 24(5), 659-685.

16.

Lord

Risi

Lambrecht

Cook

E. H.

Jr. Leventhal

B. L.

DiLavore

P. C.

Rutter

(2000). The autism diagnostic observation schedule—generic: A standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders, 30, 205-223.

17.

Macintosh

Dissanayake

(2006). Social skills and problem behaviours in school aged children with high-functioning autism and Asperger’s Disorder. Journal of Autism and Developmental Disorders, 36, 1065-1076.

18.

McConnell

S. R.

(2002). Interventions to facilitate social interaction for young children with autism: Review of available research and recommendations for educational intervention and future research. Journal of Autism and Developmental Disorders, 32, 351-372.

19.

Meier

C. R.

DiPerna

J. C.

Oster

M. M.

(2006). Importance of social skills in the elementary grades. Education & Treatment of Children, 29, 409-419.

20.

Pine

Luby

Abbacchi

Constantino

J. N.

(2006). Quantitative assessment of autistic symptomatology in preschoolers. Autism, 10, 344-352.

21.

Reisinger

E. M.

Harker

C. M.

Xie

Shin

Mandell

D. S.

(2010). Parent–teacher concordance on the Social Responsiveness Scale for children with an existing ASD diagnosis. Poster presented at the International Meeting for Autism Research, Philadelphia, PA.

22.

Rotheram-Fuller

Kasari

Chamberlain

Locke

(2010). Grade related changes in the social inclusion of children with autism in general education classrooms. Journal of Child Psychology and Psychiatry, 51, 1227-1234.

23.

Waters

Deane

K. E.

(1985). Defining and assessing individual differences in attachment relationships: Q-methodology and the organization of behavior in infancy and early childhood. Child Development, 50, 41-65.

24.

Williams White

Keonig

Scahill

(2007). Social skills development in children with autism spectrum disorders: A review of the intervention research. Journal of Autism and Developmental Disorders, 37, 1858-1868.