752098 GCQXXX10.1177/0016986217752098Gifted Child QuarterlyCederberg et al. research-article2018 Article ASD Screening Measures for High-Ability Youth With ASD: Examining the ASSQ and SRS Gifted Child Quarterly 2018, Vol. 62(2) 220–229 © 2018 National Association for Gifted Children Reprints and permissions: sagepub.com/journalsPermissions.nav https://doi.org/10.1177/0016986217752098 DOI: 10.1177/0016986217752098 journals.sagepub.com/home/gcq Charles D. Cederberg1, Lianne C. Gann1, Megan Foley-Nicpon1 , and Zachary Sussman1 Abstract High-ability youth diagnosed with autism spectrum disorder (ASD) historically have been neglected within samples validating ASD screening measures, and consensus for what constitutes high ability has not been established. The Autism Spectrum Screening Questionnaire (ASSQ) and Social Responsiveness Scale (SRS) are two common screening tools for ASD used within research and practice settings. We investigated the accuracy of the ASSQ and SRS for ASD identification among a sample of 23 high-ability youth previously diagnosed with ASD. Results suggest both ASSQ and SRS measures inconsistently screened for ASD. The high-ability students with ASD scored significantly lower on the SRS total score and social cognition, communication, and motivation subscales, suggesting potential phenotypic differences among high-ability youth with ASD that could inform assessment and intervention strategies. Keywords ASSQ, SRS, twice-exceptional, intellectual giftedness, autism spectrum disorder Children diagnosed with autism spectrum disorder (ASD) possess marked social and communication difficulties accompanied by a restricted pattern of repetitive and stereotyped behaviors (American Psychiatric Association [APA], 2013). Over the past 50 years, the process by which researchers, clinicians, and educators characterize and understand ASD has radically evolved. Early autism epidemiology studies conducted in the 1960s and 1970s approximated 2 to 5 out of 10,000 individuals met criteria for the disorder (Gillberg & Wing, 1999). Researchers during this era generally considered ASD as a condition accompanied by severe intellectual disability (Gillberg & Wing, 1999). Today, we know this condition is heterogeneously presented across individuals, and has a strong genetic component with characteristics generally identified in various stages of early childhood (Kopp & Gillberg, 2011). ASD also is observed across a wide range of cognitive abilities, from those identified with coexisting intellectual disability to those who have ability within the gifted range. Over the past several decades, rates of ASD diagnoses have substantially increased among the general population (Christensen et al., 2016). Most recent estimates suggest an overall prevalence of ASD to be approximately 1 in 68 children, with an estimated half of which possess intellectual abilities within the average to above average range (Christensen et al., 2016). Despite these diverging variants among ASD characteristics, there exists support for autistic traits sharing a unitary underlying factor structure that can be measured with appropriate tools (Constantino et al., 2004). Researchers have increasingly investigated the high-ability subset of children who meet criteria for both ASD and intellectual giftedness (Assouline, Nicpon, & Doobay, 2009). Gifted youth with ASD often show high intellect accompanied by intense interests, asynchronous development in abilities, inability to communicate and behave in a socially conventional manner, and extreme rigidity for rules and structure (Rubenstein, Schelling, Wilczynski, & Hooks, 2015). Other characteristics include deficits in social awareness, social reciprocity, emotional regulation, recognizing the emotions of others, and sensory integration (Kuo, Liang, Tseng, & Gau, 2014; Weinfeld, Barnes-Robinson, Jeweler, & Roffman-Shevitz, 2006). Gifted youth with ASD may endorse experiencing depression and social withdrawal (Doobay, Foley-Nicpon, Ali, & Assouline, 2014); however, they often show strengths in the areas of reading, mathematics, memorization of facts, and verbal fluency (Kuo et al., 2014; Neihart, 2000; Weinfeld et al., 2006). These unique 1 University of Iowa, Iowa City, IA, USA Corresponding Author: Charles D. Cederberg, Department of Psychological and Quantitative Foundations, The University of Iowa, 361 Lindquist Center, Iowa City, IA 52242-1529, USA. Email: charles-cederberg@uiowa.edu Cederberg et al. characteristics commonly present challenges for parents and educators to provide effective support and advocacy, develop strengths, and address the individual areas of deficit for the child (Dare & Nowicki, 2015; Rubenstein et al., 2015). Within this population, the absence of a linguistic or intellectual impairment, or a complex behavioral presentation, may also make identification more difficult (Assouline et al., 2009; Neihart, 2000). There exists a need for valid and reliable screening tools for the range of autistic disorders across intellectual abilities from the general population to clinical settings (Ehlers, Gillberg, & Wing, 1999). In 2013, the publication of the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; APA, 2013) consolidated the previous diagnoses of autism disorder, pervasive developmental disorder–not otherwise specified (PDD-NOS), and Asperger’s Syndrome into one spectrum (ASD), while noting one of three possible levels of severity. This consolidation may raise challenges for mental health professionals to screen accurately across the wide phenotypical range of the ASD population. For instance, Foley-Nicpon, Fosenburg, Wurster, and Assouline (2017) found the Autism Diagnostic Observation Schedule (ADOS; Lord, Rutter, DiLavore, & Risi, 2002), a gold standard research and clinical autism instrument used for diagnostic purposes, was insufficient as a single tool for diagnosing ASD among high-ability youth under DSM-5 criteria. These results raise potential questions that other widely used ASD tools, such as screening measures, may also share similar weaknesses for positively identifying ASD among high-ability youth. Screening Measures The Social Responsiveness Scale (SRS; Constantino & Gruber, 2005) is one commonly utilized parent-completed screening tool for ASD characteristics in youth aged 4 to 18 years. The form consists of 65 items scored on a 4-point Likert-type scale, yielding a total score and five subscale scores: social awareness, social cognition, social communication, social motivation, and autistic mannerisms. In 2012, the Social Responsiveness Scale–Second edition (SRS-2) was released as the successor to the original SRS, which was subsequently renamed the School-Age Form (Bruni, 2014). The School-Age Form remained unchanged with this transition and continued to cover age ranges 4.0 to 18.0 years. Furthermore, the SRS-2 is unaltered from the original SRS within this specified 4.0 to 18.0 age range (Frazier et al., 2014). The main addition to the SRS-2 provided extended age ranges up through adulthood and down through preschool. These extended age groups benefit from the inclusion of separate versions using new forms for respective preschool and adult age groups (Bruni, 2014; Frazier et al., 2014). The SRS-2 also provides more detailed qualitative descriptors in delineating regions of clinical T-scores by separating “mild” and “moderate” into two categorically distinct 221 regions (Constantino & Gruber, 2012). Clinical T-scores also have been renormed. Therefore, while a successor to the SRS might be widely available, the use of the SRS as a screening instrument remains relevant and timely. Most researchers examining the SRS have broadly claimed to include high-functioning individuals with ASD within respective study samples. However, the term high functioning encompasses an inconsistent range of abilities among research participants. This has resulted in loosely defined parameters for establishing diagnostic accuracy of the SRS. Despite these challenges, researchers using the SRS within the literature generally have included samples of children diagnosed with ASD as possessing an IQ greater or equal to 75 (e.g., Armstrong & Iarocci, 2013; Hus, Bishop, Gotham, Huerta, & Lord, 2013). This might suggest much of the research utilizing the SRS may not be applicable to children diagnosed with ASD who demonstrate intellectual giftedness with superior IQ scores (IQ > 120). This lack of inclusion of high-ability participants within samples of extant studies investigating SRS screening accuracy is not intended to call into question the measure’s utility for screening ASD among the general population. There is support for the SRS being strongly correlated with the ADOS (Lord et al., 2002) and the Childhood Autism Rating Scale (Reszka, Boyd, McBee, Hume, & Odom, 2014), for example. However, every measure possesses its own unique set of limitations. For instance, the SRS has been problematic for differential diagnostic purposes. Researchers have suggested the measure occasionally provides inaccurate diagnosis, such as mislabeling those with anxiety disorders as having ASD (Cholemkery, Mojica, Rohrmann, Gensthaler, & Freitag, 2014). Furthermore, some research suggests the SRS may overidentify children with ASD and that subscale measure scores should be interpreted with caution (Aldridge, Gibbs, Schmidhofer, & Williams, 2012). Additionally, there may be disagreement with new DSM-5 symptom criteria and the SRS, especially for children (Reszka et al., 2014). The Autism Spectrum Screening Questionnaire (ASSQ; Ehlers et al., 1999) is another popular screening instrument found to be reliable and valid in identifying a three-factored structure of ASD characteristics in school-aged children (Ryland, Hysing, Posserud, Gillberg, & Lundervold, 2012). The ASSQ is purported to be valuable particularly in singling out children in need of further assessment for an ASD diagnosis (Posserud, Lundervold, & Gillberg, 2006). Results from a factor analysis suggest that ASSQ items diverge into a three-factor model: social difficulties, tics/motor/obsessive–compulsive disorder, and autistic style (Posserud, Lundervold, & Gillberg, 2009). Similar to the SRS, the ASSQ is intended to screen for ASD among high-functioning youth. However, researchers who use the ASSQ have inconsistently established lower limit IQ cutoff scores across study samples. Posserud et al. (2009) found the ASSQ to be effective in screening ASD in typical and low-functioning children, though no official 222 cutoff IQ score was reported outside of generalizing the study’s findings to children with average intelligence to mild intellectual disability (IQ > 50). Mattila et al. (2009) also employed the IQ > 50 cutoff score in labeling “high-functioning” children, and recommended the questionnaire be used with children of average intelligence or mild intellectual disability. Ehlers et al. (1999) included children with mild intellectual disability within their high-functioning sample due to observed similarities in presentation between these children and children with higher IQs, although no cutoff score was reported. In a recent large validation study of the ASSQ for both clinical settings and the general population in Finland, Mattila et al. (2012) established score tiers of risk for ASD for children aged 7 to 12 years. In balancing specificity and selectivity, the authors established combined teacher and parent rating scores of 30 to meet criteria for medium risk of ASD diagnosis. However, while Mattila et al. (2012) were able to establish a cutoff of 22 in clinical settings for teacher ratings, they were unable to determine an optimal cutoff for parent ASSQ ratings alone. A further concern regards the lack of ASSQ sensitivity for identifying girls who meet criteria for ASD (Posserud et al., 2009). Specifically, Posserud et al. (2009) found that parent and teacher ASSQ scores were particularly discrepant for girls in a large sample of students. Kopp and Gillberg (2011) later revised the ASSQ to accommodate for this discrepancy with the aim of increased sensitivity to the presentation of ASD characteristics within female child populations. These modifications namely included the addition of 18 items conceptualized to better identify girls with ASD. Revisions also identified a flexible cutoff score of IQ greater or equal to 70. Consistently in the literature, the ASSQ is found to be useful in screening for ASD characteristics with a wide range of intellectual abilities; however, little to no research has investigated its sensitivity and utility for screening ASD in intellectually gifted children with IQ > 120. Additionally, the broad cognitive diversity commonly used in ASSQ study samples may hinder the identification of individuals with higher cognitive functioning. The current study examines the use of the SRS and ASSQ among a sample of high-ability children with preexisting diagnoses of DSM-IV-TR (APA, 2000) developmental disorders now subsumed under the ASD umbrella in the DSM-5 (APA, 2013). We are specifically interested in learning how high-ability students diagnosed with ASD phenotypically compare on SRS and ASSQ measures with non-high-ability ASD clinical samples. The following research questions were asked: Research Question 1: How do the SRS and ASSQ compare in confirming diagnoses of ASD among previously diagnosed high-ability students with ASD? Research Question 2: How do SRS scores among the high-ability sample compare with scores from a clinical sample of non-high-ability children with ASD? Gifted Child Quarterly 62(2) Method Measures The SRS (Constantino & Gruber, 2005) is a quantitative measure consisting of 65 items scored on a 4-point Likerttype scale using a “0” (not true) to “3” (almost always true). Questions can be completed in 15 to 20 minutes and focus on the child’s ability to engage in emotionally appropriate, reciprocal social interactions during the past 6 months. It is a parent- and/or teacher-report measure that assesses the child’s capacity to engage in reciprocal social behavior, interpret the emotional and interpersonal cues of others, accurately interpret these cues, appropriately respond to social cue interpretations, and possess motivation to seek out social interactions with others (Constantino et al., 2003). The measure provides a total score and five subscale scores, including social awareness (e.g., understanding personal boundaries of others or invading someone’s space), social cognition (e.g., focusing disproportionately on “parts” of things rather than “seeing the whole picture), social communication (e.g., demonstrating rigid or inflexible patterns of behavior), social motivation (e.g., seeking out social interactions on one’s own volition), and autistic mannerisms (e.g., demonstrates odd, repetitive behaviors, such as rocking or hand flapping). These raw scores are converted into T-scores that in turn provide confidence indexes for different diagnostic categories (e.g., Asperger’s Syndrome, autistic disorder, and PDD-NOS). The SRS authors report an overall clinical internal consistency of 0.97 and a parent rating Cronbach’s alpha score of .945 (Constantino & Gruber, 2005). The ASSQ (Ehlers et al., 1999) can be completed in approximately 10 minutes, and screens for possible ASD characteristics in children and adolescent populations possessing higher intellectual functioning. The ASSQ is a parent- and/or teacher-report form consisting of 27 items that requires the reporter to indicate for each item if “the rated child stands out as different from other children of his or her age.” A 3-point scale is used that ranges from No (Score 0 indicating normal), Somewhat (Score 1 indicating some abnormality), or Yes (Score 2 indicating definite abnormality; Mattila et al., 2012). The ASSQ scores ranges from 0 to 54, and include cutoff scores of 19 for parent reporters and 22 for teacher reporters to identify the likely presence of high-functioning ASD in clinical settings. The ASSQ is designed to measure three factors, including “social difficulties” (e.g., difficulties forming friendships, engaging in prosocial behaviors, and social communication), “repetitive, stereotyped behavior, and autism-associated problems” (e.g., motor difficulties, tics), and “autistic style” (e.g., social–cognitive speaking and mannerisms often encountered in highfunctioning individuals with ASD). Research has shown the ASSQ to be an accurate screening measure, with good reliability, validity, and internal consistency (Posserud et al., 2009) for individuals with low to average intelligence (Ehlers et al., 1999) across clinical 223 Cederberg et al. settings and the general population (Posserud et al., 2009). Ehlers et al. (1999) reported high test–retest reliability (r = .94 for parents, r = .96 for teachers). Interrater reliability between parents and teachers was .66. Divergent validity between parent ratings on the ASSQ and the Rutters and 10-item Conners scales, two widely used measures for assessing behavioral disorders, were .75 to .58, respectively (Ehlers et al., 1999). While the authors conceded that the agreement between the ASSQ and Rutters scales was somewhat high, they attributed this to the relatively broad range of behavioral characteristics captured by the Rutters scale, some of which reflect features of ASD. Table 1. Means and Standard Deviations for Study Sample Participants’ Ability Scores. Index Mean Standard deviation Full-scale General ability index Verbal comprehension index Perceptual reasoning index 140.00 138.33 123.15 122.58 6.14 10.02 15.39 13.59 Participants included a total of 23 parents whose children and adolescents were identified with intellectual giftedness (IQ score of 120 [93rd percentile] or above) and met diagnostic criteria for an ASD. Age ranges of child and adolescent participants were 4.0 years to 17.11 years, with a mean age of 11.93 years and standard deviation of 3.43 years. 6.1 to 16.0 years, and Wechsler Adult Intelligence Scale– Fourth edition (Wechsler, 2008) for adolescents older than 16.0 years, of the sample. ASD diagnosis was determined through administration of a comprehensive evaluation, including the ADOS (Lord et al., 2002), the Autism Diagnostic Interview–Revised (ADI-R; Rutter, Le Couteur, & Lord, 2003), and the Vineland Adaptive Behavior Scales–Second edition (Sparrow, Cicchetti, & Balla, 2005). Diagnoses were made by one of two licensed psychologists who had received research training in administering the ADOS and ADI-R. Procedures Statistical Analysis Participants were recruited from two separate avenues. First, a grant-funded study provided the opportunity for psychologists with requisite clinical training to assess students with highintellectual ability accompanied by coexisting ASD diagnoses. Second, additional students were recruited through a Midwestern university-based child and adolescent psychology clinic. These student diagnostic profiles were consistent with high ability and an ASD diagnosis. Parents were initially contacted via e-mail with information regarding the study. Interested participants were then mailed a packet with a cover letter explaining the purpose of the study and instructions for completing the assessments (order of completion was not specified), consent and assent forms, the SRS and ASSQ, and a return envelope. Individuals were asked to return all documentation (informed consent documents and study materials) in the supplied envelope. If participants did not return the envelope in 4 weeks, a reminder e-mail was sent. If a participant returned a packet with an incomplete informed consent document, he or she was mailed the informed consent document a second time with a brief note indicating where a signature was required. If someone returned the informed consent document, but not the study materials, he or she was sent a brief e-mail reminder to return the study materials when completed. Intellectual ability for giftedness was determined through administering Wechsler intelligence testing. Table 1 provides the means and standard deviations of the age appropriate Wechsler test, Wechsler Preschool and Primary Scale of Intelligence–Third edition (Wechsler, 2002) for children up to age 6.0 years; Wechsler Intelligence Scale for Children–Fourth edition (Wechsler, 2004) for children We ran a cross-tabulation analysis for both ASSQ and SRS measures with predetermined cutoff scores of 19 and 85, respectively. This analysis served to delineate which previously diagnosed ASD participant scores met inclusion for ASD based on performance of the ASSQ or the SRS. Finally, two-tailed, one sample independent t tests were used to compare mean total scores and subscale (i.e., social awareness, social cognition, social communication, social motivation, and autistic mannerisms) scores from the high-ability sample with the SRS sample total and subscale scores provided by Cholemkery et al. (2014). That is, a two-tailed, one sample independent t test was conducted to determine whether differences exist between the gifted and typical ASD clinical samples for the total raw score and each of the five raw subscale scores. Significance level was set at α = .05 (uncorrected). Participants Results We compared the ASSQ and SRS as tools in confirming diagnoses of ASD among previously diagnosed high-ability youth with ASD. We used cross-tabulation to examine the effectiveness of confirming ASD diagnosis with each measure. Additionally, participant ASSQ and SRS scores were entered in a cross-tabulation analysis after transforming the raw scores from the ASSQ and SRS into categorical data (i.e., diagnosis). The authors utilized a cutoff score of 19 for the ASSQ, so positive cases of ASD were defined as a score of 19 or higher. This cutoff score was determined by a sensitivity and specificity receiver operating characteristic curve analysis conducted by Ehlers et al. (1999). Ehlers et al. (1999) established this optimal cutoff score differentiating 224 Gifted Child Quarterly 62(2) Table 2. ASSQ and SRS Diagnosis Cross-Tabulation Analysis. SRS ASSQ No dx Count % Within ASSQ Yes dx Count % Within Total Count % Within No dx Yes dx Total 8 100 0 0 8 100 5 33.3 10 66.7 15 100 13 56.5 10 43.5 23 100 Note. ASSQ = Autism Spectrum Screening Questionnaire; SRS = Social Responsiveness Scale. ASD cases from other cases to be 19 for parent responders, yielding an 82% correct identification of Asperger syndrome in a validation sample. The authors utilized a cutoff score of 85 for the SRS based on the clinical cutoff recommendation in the SRS and SRS-2 manuals (Constantino & Gruber, 2005, 2012). The ASSQ and SRS scores positively agreed with ASD diagnosis across 10 total cases (43.5% of the total sample). Individually, the ASSQ identified more positive ASD diagnoses (n = 15) than the SRS (n = 10; see Table 2). Furthermore, no SRS scores were found to positively identify ASD among participants who scored below the established ASSQ cutoff score. While both tests were associated with each other (K = .582; p = .002), we suggest these findings indicate that the ASSQ may possess higher sensitivity for correctly identifying ASD among high-ability youth as a single-screening measure. Conversely, the SRS appears to be more restrictive in positively identifying ASD among highability cases. Considering all cases were previously diagnosed with ASD by a licensed psychologist (n = 23), the ASSQ and SRS might also be limited in their ability to screen for ASD among high-ability individuals when used in tandem. In our second research question, we investigated how SRS scores among the high-ability sample compared with scores from a clinical sample of non-high-ability children with ASD. We conducted one sample t tests of SRS subscale mean scores in comparison with respective subscale mean scores from the ASD clinical sample provided in Cholemkery et al. (2014; see Table 3). The Cholemkery et al.’s study is to our knowledge the only recently published article that provided descriptive statistics for subscale SRS scores and total raw scores for a clinical ASD sample; no subscale-level scores were provided by Constantino and Gruber (2005). We also verified average intellectual Full Scale Intelligence Quotient ability in the Cholemkery et al. study to ensure it was composed of an ASD nongifted clinical sample (M = 102.15; SD = 16.23). We observed significantly lower scores within our high-ability sample on the cognition (p = .046; d = .446; 95% confidence interval [CI: −5.641, −0.49]), communication (p = .016; d = .556; 95% CI [−9.813, −1.13]), and motivation (p = .001; d = .785; 95% CI [−6.546, −1.954]) SRS subscales. No significant differences were found between the high ability and Cholemkery et al. groups for the awareness (p = .287; d = .221; 95% CI [−2.093, 0.651]) and autistic mannerisms (p = .795; d = .059; 95% CI [−2.737, 3.532]) subscales. Additionally, the total raw score for the high-ability sample was significantly lower than the Cholemkery et al. total raw score (p = .033; d = .484; 95% CI [−24.612, −1.17]). Discussion In the current study, we examined the use of two ASD screening tools among high-ability populations suspected of ASD. Specifically, we examined whether the ASSQ and SRS confirmed an ASD diagnosis among 23 high-ability youth previously diagnosed with ASD. We then investigated how the SRS measure scores among a high-ability sample compare with scores from a clinical sample of non-high-ability children with ASD. The goal was to determine the adequacy of each screening measure and explore potential presentation differences for clinicians and researchers to consider when assessing for ASD in high-ability individuals. While these findings are preliminary, the results yielded a number of interesting findings that warrant consideration of the ASSQ and SRS screening measures’ utility in accurately identifying high-ability youth with ASD. A licensed psychologist using gold standard ASD identification instruments prior to involvement in the study diagnosed all participants with ASD. Despite the utility of the SRS and ASSQ as screeners, only a small percentage of cases in the current sample were identified by both measures. In comparing the SRS and ASSQ, the cross-tabulation analysis demonstrated that the ASSQ possessed a high degree of sensitivity for positively identifying ASD. In contrast, the SRS was more restrictive in identifying ASD compared to the ASSQ. Therefore, it could be suggested that the ASSQ was more fitting to identify ASD in a high-ability population because it had a lower threshold and identified more potential cases of ASD. Despite the ASSQ’s lower threshold, both measures yielded inconsistencies as screening measures as evidenced by the ASSQ and SRS positively identifying ASD in only 15 and 10 cases out of the total 23 participants, respectively. Moreover, only 43.5% of the cases in the current sample were positively identified with ASD by both screening measures. This presents a challenge because the purpose of screeners is to balance a variety of costs and benefits for specificity and sensitivity in hopes of casting a wide net for potential symptomology (e.g., Campbell, 2005; Mattila et al., 2012). For example, the authors of the SRS screener describe its purpose as to identify subthreshold abilities and deficits in social reciprocity that may be indicative of a pervasive development disorder (Constantino & Gruber, 2005). However, if the cutoff score for screeners such as the 225 Cederberg et al. Table 3. One Sample t Test SRS Total and Subscale Scores Between ASD Gifted and Cholemkery et al. (2014) Normal Cognitive Ability ASD Groups. Total score Social awareness Social cognition Social communication Social motivation Autistic mannerisms SRS 2E (n = 23) Cholemkery et al. (2014); ASD (n = 60) M (SD) M (SD) 85.74 (27.1) 11.61 (3.17) 15.43 (6.47) 28.48 (10.04) 13.00 (5.31) 17.22 (7.25) 98.63 (26.19) 12.33 (3.34) 18.28 (6.32) 33.95 (9.64) 17.25 (5.51) 16.82 (6.20) p .033* .287 .046* .016* .001* .795 Note. SRS = Social Responsiveness Scale; ASD = autism spectrum disorder. *Significant at the .05 level. SRS are too high, then potential cases of ASD may be missed among high-ability students. In our study, 43.5% (n = 10) of the sample scored above the recommended SRS total raw score cutoff in clinical settings (T = 85; Constantino & Gruber, 2005; 2012). Take as a whole, these findings suggest that the overall level of symptomology and/or presence of ASD diagnostic characteristics found in typical clinical samples may be more prominent than among high-ability samples. The cross tabs are a reflection of the restrictiveness of the SRS to accurately identify ASD within the high-ability sample, as 13 (56.5%) of the participants did not meet the recommended SRS clinical cutoff threshold score of 85. Our second objective of this study was to explore how the SRS measure scores among a high-ability sample compared with scores from a clinical sample of non-high-ability children with ASD. Specifically, SRS scores among the highability sample were compared with scores from a clinical sample of non-high-ability children with ASD from Cholemkery et al. (2014). The total raw SRS score of the high-ability ASD sample was significantly lower than the sample from the Cholemkery et al. (2014) study. Although preliminary and from a small sample, these results may be supportive of the “masking effect” (Assouline, Nicpon, & Whiteman, 2010), which suggests high ability may sometimes “mask” symptomology. It also may be ASD characteristics are less severe in high-ability populations, consistent with Level 1 in the DSM-5 (APA, 2013). Analysis of the SRS subscale differences between the high-ability sample and comparison sample further elucidates these differences. The SRS social awareness and social cognition subscales measure the capacity to accurately recognize, understand, and react to social cues and engage in reciprocal social interactions (Constantino & Gruber, 2005). Each subscale focuses on reciprocal social behavior, with the social awareness subscale focusing on the sensory aspects, and the social cognition subscale relating to the cognitive– interpretive aspects of the diagnosis. In the current study, there were no significant differences in social awareness between groups, suggesting the high-ability sample possessed similar deficits in social pragmatics and perspective taking. The significantly lower scores obtained among the high-ability sample on the social cognition subscale, which captures the capacity to interpret social cues and reciprocal social behavior, might be related to their higher intellectual abilities. That is, the youths’ enhanced ability to learn from feedback related to social skills might instill a greater capacity for interpreting social cues, cognitive flexibility, and adapting to norms related to social behavior. While these social deficits might still be apparent compared with youth without ASD, the results suggest they might be less so compared with average ability populations with ASD. Among the high-ability sample, social communication and social motivation subscale scores were also significantly lower than those reported in Cholemkery et al. (2014). The social communication subscale measures the expressive components of reciprocal social behavior, such as engaging in appropriate turn-taking in conversations and understanding personal space boundaries. Social motivation pertains to the extent that one is motivated to engage in interpersonal interactions, and comprises elements of social anxiety and empathy (Constantino & Gruber, 2005). These significantly lower scores observed on these subscales further supports the compensatory hypothesis. Social communication skills often can be taught or learned from a variety of interventions and naturally occurring social contexts, such as ecological variations, collateral skills interventions, child-specific interventions, and observed peer behavior (McConnell, 2002). Likewise, intellectual giftedness might facilitate more efficient and expeditious acquisition of social communication and social cognitive skills, and explain a heightened motivation for engaging in social interactions and initiating meaningful connections with peers. In sum, intellectual ability may increase one’s capacity to learn from social communication opportunities and compensate for one’s inherent social deficits. However, these hypotheses are extremely tentative given the small sample size, lack of a true comparison group, 226 and methodological limitations that will be described in the next section. It is notable there were no mean group differences between the groups on the autistic mannerisms subscale, which measures engagement in stereotypical behaviors or highly restrictive interests. Although the phenotypical pattern of symptom manifestation may be unique among high-ability youth (excessive interest in unusual or specific topics as the most frequently observed; Foley-Nicpon et al., 2017), this result suggests the overall presence of symptoms within this umbrella may be similar, regardless of cognitive ability. This finding is generally aligned to the hypotheses suggesting the existence of differential restricted and repetitive behavior subgroups in children with ASD (Bishop et al., 2013). Motor mannerisms, sensory seeking behaviors, and repetitive use of objects might be separate from an insistence on sameness. An insistence on sameness refers to cognitively rigid patterns of behavior marked by challenges with changes in routine and overreliance on rituals. Together, restricted repetitive behaviors and insistence on sameness might underpin the biological factors of ASD (Bishop et al., 2013; Neihart, 2000). One final component in considering these findings pertains to the intersection of giftedness, social development, and psychological adjustment. Generally, research suggests high-ability children possess greater confidence and overall higher self-concept than nongifted peers (Assouline & Colangelo, 2006). This follows the idea that success achieved in academic domains might be interconnected with social and emotional development (Foley-Nicpon, 2015). Building a capacity for perseverance and resiliency might be an important factor at work for some high-ability youth with ASD, and developing individual strengths might compensate for social weaknesses through trial and error learning that ultimately leads to effective social problem-solving strategies (Willard-Holt, Weber, Morrison, & Horgan, 2013). It is unclear if these processes were at play within our study, though our findings raise compelling possible avenues for future research. Relevance to the SRS-2 At the time of the study proposal and data collection, the SRS-2 had not been published. As previously described, the SRS-2 expanded the age range of the instrument into three distinct Preschool, School-Age, and Adult Forms that are used to monitor social impairment across the life span. Despite the addition of these separate Preschool and Adult Forms, the original SRS validated for youth aged 4 to 18 years was simply retermed as the School-Age Form. The 65-item rating scale of the SRS School-Age Form remains unchanged from the original SRS. However, the new SRS-2 manual provides updated norms and validation data for use in clinical and school settings. Therefore, in order to boost the utility of this study for practitioners working in these Gifted Child Quarterly 62(2) settings, we retroactively explored the possible implications to our findings using the renormed clinical T-scores of the SRS-2 School-Age Form. It is important to note that the SRS-2 manual clearly advises using SRS raw scores (i.e., possible scores varied from 1 to 195, based on 0- to 3-point Likert-type scale weighting for each of the 65 SRS test question responses; Constantino & Gruber, 2005, p. 14) for research purposes in clinical populations (Constantino & Gruber, 2012, p. 17). However, investigating clinical T-score results may inform the decision-making process of clinicians and educators who use this instrument in real-world settings. Using original SRS T-scores, we found 12 participants fell in the “severe” range, 7 fell in the “mild to moderate” range, and 4 fell in the “within normal limits” range. Using SRS-2 renormed T-scores and the more detailed qualitative descriptors, 10 participants fell in the “severe” range, 7 fell in the “moderate” range, 2 fell in the “mild” range, and 4 were identified as “within normal limits.” Therefore, two participants moved from “severe” to “moderate” using renormed SRS-2 T-scores, while no new participants fell “within normal limits.” The SRS-2 manual outlines detailed interpretative text for each of the different SRS T-score ranges. Those scores falling within the “severe” range are highly associated with a clinical diagnosis of ASD, while scores within the “moderate” range are common among children with ASD who demonstrate substantial deficits with everyday social interactions (Constantino & Gruber, 2012). These represented 17 of the 23 participants in our study. Moreover, scores falling in the “mild” range warrant careful clinical judgment, as these scores are also commonly observed across a variety of non-ASD diagnoses such as attention-deficit/hyperactivity disorder, obsessive–compulsive disorder, anxiety disorders, conduct disorders, and the new DSM-5 social communication disorder. Considering all 23 of our participants began the study with a diagnosis of ASD, this wide diversity of observed SRS-2 scores strengthens our findings that a high degree of care is needed among clinicians and educators when recommending ASD evaluations to high-ability children. The variety of T-scores also supports previous findings indicating ASD manifests distinctively among high-ability youth with ASD (Doobay et al., 2014; Foley-Nicpon et al., 2010; Foley-Nicpon et al., 2017; Neihart, 2000). These findings are not meant to be an indictment of the SRS-2 as a tool for screening ASD among high-ability youth. The SRS-2 is specifically designed to assess for autistic symptoms rapidly and reliably within a proper context, and it also explicitly avoids “all or nothing” categorical diagnoses (Constantino & Gruber, 2012). However, because ASD in high-ability youth is so widely misunderstood among educators and the lay public (e.g., Dare & Nowicki, 2015; Rubenstein et al., 2015), interpreting SRS-2 scores within the context of a child’s ability is especially important. The addition of social communication disorder in the DSM-5 227 Cederberg et al. further complicates how practitioners in school and clinical settings interpret scores from widely established ASD instruments (Foley-Nicpon et al., 2017), including the SRS-2. Had our SRS-2 data been used to recommend comprehensive ASD evaluations in the real world, it is quite possible 6 of our 23 participants with ASD may never have received further ASD diagnostic services. authors’ knowledge, the ASSQ is not normed (Campbell, 2005). This prohibits the availability of floor, ceiling, and item gradient information. Despite these limitations, the results of the study provide initial insight into the unique symptom presentation of high-ability children with ASD and the utility of ASD screeners with this specific population. Limitations Conclusion There are several limitations to this study that warrant thoughtful discourse. First, the absence of a comparison group consisting of children of average cognitive abilities with ASD, or high-ability children without ASD, limited the opportunity for alternative methodological approaches. Second, participants in the current study were diagnosed with ASD based on DSM-IV-TR criteria for autistic disorder, Asperger syndrome, or PDD-NOS. It is not certain how many of the participants would meet criteria for ASD under the current DSM-5 diagnostic criteria, though Frazier et al. (2012) found support for the validity of DSM-5 criteria for ASD and for the continued utility of the SRS-2 as a screening measure. Frazier et al. (2012) identified a two-factor model composed of autistic mannerisms and social communication that later informed the DSM-5 taxonomy. Additionally, though specificity of the DSM-5 was found to be higher for ASD diagnosis, sensitivity was lower when compared with DSM-IV-TR criteria. A third limitation is that comparison data derived from the selected Cholemkery et al. (2013) study was not normed and possesses its own set of methodological issues. This study was conducted with a sample from Germany, so crosscultural differences in scores might confound results. The choice to use the Cholemkery et al. (2013) article was due to the unavailability of subscale clinical means for ASD samples within the SRS and SRS-2 manuals (Constantino & Gruber, 2005). Therefore, we could not draw on such data for the purposes of comparative analysis. Fourth, this study possesses a small sample size and thus generalizability is hindered. However, sample size problems in research are commonly encountered issues among low incidence populations, such as high-ability children with ASD. Considering the small sample size and challenges with statistical power, utilizing a t test implies a high probability of committing false positive Type II error (e.g., Cohen, 1970). Fifth, the participant age range of 4 to 18 years was large, but this range is within the range recommended for use of the SRS and the SRS-2 School-Age Form instruments (Constantino & Gruber, 2012). Sixth, there are many different measures available to screen for ASD, and the SRS and ASSQ may be less effective in identifying high-ability students with ASD than other existing measures that were not included within the scope of this study. However, the SRS and ASSQ were chosen because both measures claim to screen for ASD in high-functioning individuals (>70 IQ). Finally, to the High-ability youth with ASD may present differently than ASD youth without high ability, yet the majority of the measures used to screen ASD populations are created with samples of individuals with low average to average intelligence. This presents problems when considering the unique social challenges encountered by high-ability youth with ASD. Future screening instruments should consider the social communication, social motivation, and social cognition characteristics that potentially distinguish high-ability children with ASD from other groups of youth with ASD. Screening instruments should be developed that accurately identify and meaningfully capture these phenotypic differences to avoid missed diagnosis among high-ability children with ASD. 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 The author(s) received no financial support for the research, authorship, and/or publication of this article. ORCID iD Megan Foley-Nicpon https://orcid.org/0000-0002-8677-3578 References Aldridge, F. J., Gibbs, V. M., Schmidhofer, K., & Williams, M. (2012). Investigating the clinical usefulness of the Social Responsiveness Scale (SRS) in a tertiary level, autism spectrum disorder specific assessment clinic. Journal of Autism and Developmental Disorders, 42, 294-300. doi:10.1007/s10803011-1242-9 American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Armstrong, K., & Iarocci, G. (2013). Brief report: The autism spectrum quotient has convergent validity with the Social Responsiveness Scale in a high-functioning sample. Journal of Autism and Developmental Disorders, 43, 2228-2232. doi:10.1007/s10803-013-1769-z 228 Assouline, S. G., & Colangelo, N. (2006). Social-emotional development of gifted adolescents. In F. A. Dixon & S. M. Moon (Eds.), The handbook of secondary gifted education (pp. 6595). Waco, TX: Prufrock Press. Assouline, S. G., Nicpon, M. F., & Doobay, A. (2009). Profoundly gifted girls and autism spectrum disorder: A psychometric case study comparison. Gifted Child Quarterly, 53, 89-105. doi:10.1177/0016986208330565 Assouline, S. G., Nicpon, M. F., & Whiteman, C. (2010). Cognitive and psychosocial characteristics of gifted students with written language disability. Gifted Child Quarterly, 54, 102-115. doi:10.1177/0016986209355974 Bishop, S. L., Hus, V., Duncan, A., Huerta, M., Gotham, K., Pickles, A., . . . Lord, C. (2013). Subcategories of restricted and repetitive behaviors in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 32, 1287-1297. doi:10.1007/s10803-012-1671-0 Bruni, T. P. (2014). Test review: Social Responsiveness Scale– Second Edition (SRS-2). Journal of Psychoeducational Assessment, 32, 365-369. doi:10.1177/0734282913517525 Campbell, J. M. (2005). Diagnostic assessment of Asperger’s disorder: A review of five third-party rating scales. Journal of Autism and Developmental Disorders, 35, 25-35. doi:10.1007/ s10803-004-1028-4 Cholemkery, H., Mojica, L., Rohrmann, S., Gensthaler, A., & Freitag, C. (2014). Can autism spectrum disorders and social anxiety disorders be differentiated by the Social Responsiveness Scale in children and adolescents? Journal of Autism and Developmental Disorders, 44, 1168-1182. doi:10.1007/s10803-013-1979-4 Christensen, D. L., Baio, J., Braun, K. V., Bildre, D., Charles, J., Constantino, J. N., . . . Yeargin-Allsopp, M. (2016). Prevalence and characteristics of autism spectrum disorder among children aged 8 years—Autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2012. Morbidity and Mortality Weekly Report (MMWR), 65(3), 1-23. doi:10.15585/mmwr.ss6503a1 Cohen, J. (1970). Approximate power and sample size determination for common one-sample and two-sample hypothesis tests. Educational and Psychological Measurement, 30, 811-831. doi:10.1177/001316447003000404 Constantino, J. N., Davis, S. A., Todd, R. D., Schindler, M. K., Gross, M. M., Brophy, S. L., . . . Reich, W. (2003). Validation of a brief quantitative measure of autistic traits: Comparison of the Social Responsiveness Scale with the Autism Diagnostic Interview-Revised. Journal of Autism and Developmental Disorders, 33, 427-433. doi:10.1023/A:1025014929212 Constantino, J. N., & Gruber, C. P. (2005). The Social Responsiveness Scale: Manual. Los Angeles, CA: Western Psychological Services. Constantino, J. N., & Gruber, C. P. (2012). Social Responsiveness Scale, second edition (SRS-2). Torrance, CA: Western Psychological Services. Constantino, J. N., Gruber, C. P., Davis, S., Hayes, S., Passanante, N., & Przybeck, T. (2004). The factor structure of autistic traits. Journal of Child Psychology and Psychiatry, 45, 719726. doi:10.1111/j.1469-7610.2004.00266.x Dare, L., & Nowicki, E. A. (2015). Twice-exceptionality: Parents’ perspectives on 2e identification, Roeper Review, 37, 208-218. doi:10.1080/02783193.2015.1077911 Gifted Child Quarterly 62(2) Doobay, A. F., Foley-Nicpon, M., Ali, S., & Assouline, S. G. (2014). Cognitive, adaptive and psychosocial differences between high ability youth with and without autism spectrum disorders. Journal of Autism and Developmental Disorders, 44, 2026-2040. doi:10.1007/s10803-014-2082-1 Ehlers, S., Gillberg, C., & Wing, L. (1999). A screening questionnaire for Asperger syndrome and other high-functioning autism spectrum disorders in school age children. Journal of Autism and Developmental Disorders, 29, 129-141. doi:10.1023/ A:1023040610384 Foley-Nicpon, M. (2015). The social and emotional development of twice-exceptional children. In M. Neihart, S. Pfeiffer, & T. Cross (Eds.), Social and emotional development of gifted children: What do we know? (pp. 103-118). Waco, TX: Prufrock Press. Foley-Nicpon, M., Doobay, A., & Assouline, S. G. (2010). Parent, teacher, and self perceptions of psychosocial functioning in intellectually gifted children and adolescents with autism spectrum disorder. Journal of Autism and Developmental Disabilities, 40, 1028-1038. doi:10.1007/s10803-010-0952-8 Foley-Nicpon, M., Fosenburg, S. L., Wurster, K. G., & Assouline, S. G. (2017). Identifying high ability children with DSM-5 autism spectrum or social communication disorder: Performance on autism diagnostic instruments. Journal of Autism and Developmental Disabilities, 47, 460-471. doi:10.1007/s10803016-2973-4 Frazier, T. W., Ratliff, K. R., Gruber, C., Zhang, Y., Law, P. A., & Constantino, J. N. (2014). Confirmatory factor analytic structure and measurement invariance of quantitative autistic traits measured by the Social Responsiveness Scale-2. Autism, 18, 31-44. doi:10.1177/1362361313500382 Frazier, T. W., Youngstrom, E. A., Speer, L., Embacher, R., Law, P., Constantino, J., . . . Eng, C. (2012). Validation of proposed DSM-5 criteria for autism spectrum disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 51, 2840. doi:10.1016/j.jaac.2011.09.021 Gillberg, C., & Wing, L. (1999). Autism: Not an extremely rare disorder. Acta Psychiatrica Scandinavica, 99, 399-406. doi:10.1111/ j.1600-0447.1999.tb00984.x Hus, V., Bishop, S., Gotham, K., Huerta, M., & Lord, C. (2013). Factors influencing scores on the Social Responsiveness Scale. Journal of Child Psychology and Psychiatry, 54, 216-224. doi:10.1111/j.1469-7610.2012.02589.x Kopp, S., & Gillberg, C. (2011). The Autism Spectrum Screening Questionnaire (ASSQ)-Revised Extended Version (ASSQREV): An instrument for better capturing the autism phenotype in girls? A preliminary study involving 191 clinical cases and community controls. Research in Developmental Disabilities, 32, 2875-2888. doi:10.1016/j.ridd.2011.05.017 Kuo, C. C., Liang, K. C., Tseng, C. C., & Gau, S. S. F. (2014). Comparison of the cognitive profiles and social adjustment between mathematically and scientifically talented students with Asperger syndrome. Research in Autism Spectrum Disorders, 8, 838-850. doi:10.1016/j.rasd.2014.04.004 Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (2002). Autism diagnostic observation schedule. Los Angeles, CA: Western Psychological Services. Mattila, M. L., Jussila, K., Kuusikko, S., Kielinen, M., Linna, S. L., Ebeling, H., . . . Moilanen, I. (2009). When does the Autism Spectrum Screening Questionnaire (ASSQ) predict autism spectrum disorders 229 Cederberg et al. in primary school-aged children? European Child & Adolescent Psychiatry, 18, 499-509. doi:10.1007/s00787-009-0044-5 Mattila, M. L., Jussila, K., Linna, S. L., Kielinen, M., Bloigu, R., Kuusikko-Gauffin, S., . . . Moilanen, I. (2012). Validation of the Finnish Autism Spectrum Screening Questionnaire (ASSQ) for clinical settings and total population screening. Journal of Autism and Developmental Disorders, 42, 2162-2180. doi:10.1007/s10803-012-1464-5 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. doi:10.1023/A:1020537805154 Neihart, M. (2000). Gifted children with Asperger’s syndrome. Gifted Child Quarterly, 44, 222-230. doi:10.1177/0016986 20004400403 Posserud, M. B., Lundervold, A. J., & Gillberg, C. (2006). Autistic features in a total population of 7–9-year-old children assessed by the ASSQ (Autism Spectrum Screening Questionnaire). Journal of Child Psychology and Psychiatry, 47, 167-175. doi:10.1111/j.1469-7610.2005.01462.x Posserud, M. B., Lundervold, A. J., & Gillberg, C. (2009). Validation of the Autism Spectrum Screening Questionnaire in a total population sample. Journal of Autism and Developmental Disorders, 39, 126-134. doi:10.1007/s10803-008-0609-z Reszka, S. S., Boyd, B. A., McBee, M., Hume, K. A., & Odom, S. L. (2014). Brief report: Concurrent validity of autism symptom severity measures. Journal of Autism and Developmental Disorders, 44, 466-470. doi:10.1007/s10803-013-1879-7 Rubenstein, L. D., Schelling, N., Wilczynski, S. M., & Hooks, E. N. (2015). Lived experiences of parents of gifted students with autism spectrum disorder: The struggles to find appropriate educational experiences. Gifted Child Quarterly, 59, 283-298. doi:10.1177/0016986215592193 Rutter, M., Le Couteur, A., & Lord, C. (2003). Autism diagnostic interview-revised. Los Angeles, CA: Western Psychological Services. Ryland, H. K., Hysing, M., Posserud, M. B., Gillberg, C., & Lundervold, A. J. (2012). Autism spectrum in children with neurological disorders. Child and Adolescent Psychiatry and Mental Health, 6(1), 1-9. doi:10.1186/1753-2000-6-34 Sparrow, S., Cicchetti, D., & Balla, D. (2005). Vineland adaptive behavior scales (2nd ed.). Minneapolis, MN: Pearson Assessments. Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence (3rd ed.). San Antonio, TX: Psychological Corporation. Wechsler, D. (2004). Wechsler Intelligence Scale for Children (4th ed.). San Antonio, TX: Psychological Corporation. Wechsler, D. (2008). Wechsler Adult Intelligence Scale (4th ed.). San Antonio, TX: Psychological Corporation. Weinfeld, R., Barnes-Robinson, L., Jeweler, S., & RoffmanShevitz, B. (2006). Smart kids with learning difficulties: Overcoming obstacles and realizing potential. Waco, TX: Prufrock Press. Willard-Holt, C., Weber, J., Morrison, K. L., & Horgan, J. (2013). Twice-exceptional learners’ perspectives on effective learning strategies. Gifted Child Quarterly, 57, 247-262. doi:10.1177/0016986213501076 Author Biographies Charles D. Cederberg received his MA in mental health counseling from Boston College and is currently completing a PhD in counseling psychology in the Department of Psychological and Quantitative Foundations at The University of Iowa. His areas of professional interest include psychosocial well-being and talent development of high-ability youth and twice-exceptional youth; social class and classism; materialism, greed, and mental health; and personal branding in professional psychology. Lianne C. Gann received her MA in clinical mental health counseling and is currently a doctoral candidate in counseling psychology at The University of Iowa. Her areas of professional interest include eating disorders, social media’s impact on eating and health behaviors, and body image. Megan Foley-Nicpon is an associate professor of counseling psychology and associate director for research and clinic at the Belin-Blank Center for Gifted Education and Talent Development at The University of Iowa. She is a licensed psychologist whose research and clinical interests include assessment and intervention with high-ability students with disabilities, and the social and emotional development of talented and diverse students. Zachary Sussman is a pediatric neuropsychologist and licensed psychologist in private practice in Boulder, Colorado. He obtained his Ph.D. in counseilng psychology from the University of Iowa.