Validity of a Brief Joint Attention Scale Based On Items From the SCQ and SRS

Background:   Joint Attention deficits are a hallmark of the social-communication symptoms of Autism Spectrum Disorders (ASD). Because of its central role in the neurodevelopmental disturbance of ASD, it is important to develop brief and accurate measures of Joint Attention (JA) to facilitate research on the genetics, broad phenotype and early identification of this disorder (Mundy, Sullivan, & Mastergeorge, 2009). Recent, efforts in this regard has led to the development of a 5-item behavioral observation measure of JA for early identification in a large sample of toddlers (Nygren, Sandberg, Gillstedt, Ekeroth, Arvidsson, & Gillberg, 2012), and a study of genetics using a JA factor isolated from the Autism Diagnostic Interview-Revised (Liu, et al., 2011). 

Objectives: The goal of this study was to determine if a valid, brief and theoretically grounded JA scale could be derived from parent reports of ASD symptomatology on the Social Communication Scale (SCQ; Rutter, Bailey, Berument, Lord, & Pickles, 2003) and the Social Responsiveness Scale (SRS; Constantino, et al., 2003). These scales were chosen because they are readily available to researchers and can be used efficiently in large scale genetic and epidemiological studies.   

Methods:   Participants included 51 typically developing children (M = 11.88 years, SD = 2.47) and 34 children with a community diagnosis of ASD (M = 11.85 years, SD = 2.52). All participants had an IQ greater than 75 on the Wechsler Abbreviated Scale of Intelligence (WASI, Wechsler, 1999), with a mean Full Scale IQ of 108.03 (16.58) and 113.91 (14.84) for the ASD and TD samples, respectively. Parents of the participants completed the SRS and the SCQ. Items addressing JA were selected from both questionnaires to comprise a JA scale. Joint attention items were selected on the basis that they assessed the tendency to engage in behaviors that involved adopting a common point of view or point of reference, either visual or mental, with another person. Seventeen items were selected from the SCQ and SRS and an iterative process of item elimination, based on item-to-item covariance, yielded an 11-item scale (7 SCQ and 4 SRS) that had optimal face validity and psychometric properties.  

Results:  The SCQ-SRS JA scale had acceptable internal consistency (α = .836). A principal components factor analysis indicated that the factor structure of the scale reflected 3 domains: Eye Contact & Social Cognition (ECSC); Joint Visual Attention (JVA); and Social-Communication Gestures (SCG). Discriminant function analysis revealed that the first two factors distinguished higher function children with autism from controls (N = 85) with 88.2% sensitivity and 90.2% specificity. The JVA factor was also significantly correlated with social orienting to peripheral peer avatars in a virtual reality social-attention public speaking task (r = -.383, p < .05 and r = -.408, p < .05). 

Conclusions:   The results of this study indicate that a valid and useful brief JA parent report measure may be obtained from the SCQ and SRS. These data suggest that additional research and development of a SCQ-SRS JA scale is warranted in studies of larger samples.