Utility of the Screening Tool for Autism in Two-Year Olds (STAT) as a Continuous Measure of Nonverbal Social-Communicative Behavior

Background: The STAT is an autism screening tool designed for children between 24 and 36 months old based on direct measurement of play, imitation and communication behavior (Stone et al., 2000; 2004).  Its screening properties have been described for children as young as 14 months (Stone et al., 2008). Because the STAT was designed to minimize language demands while measuring social-communicative behaviors that are disrupted in autism, the Total STAT score may be useful as a measure of nonverbal social-communication development in young children at risk for autism.  

Objectives: To explore the validity of the STAT as a measure nonverbal social-communicative skill in toddlers with early autism symptoms by examining its relation to social-communicative items from parent report and other observational measures.

Methods: Participants included 48 toddlers (mean CA = 21.5 months, range = 15.5–25 months) from the initial assessment of a multi-site clinical randomized trial of the Hanen More than Words Intervention who met a predetermined cutoff score on the STAT and had a clinical presentation consistent with ASD.  Pearson correlations between the Total STAT score (which ranges from 0-4, with higher scores indicating greater autism symptomatology) and related measures of social-communicative behaviors were examined.  Observational measures were the Early Social Communication Scale (ESCS) initiating joint attention (IJA), responding to joint attention (RJA) and initiating behavioral requests (IBR) scores and the Developmental Play Assessment (DPA) score for number of toys used during a 7-minute period. Parental report measures were the Parent Interview for Autism-Clinical Version (PIA-CV) Imitation and Nonverbal Communication scores and the Infant-Toddler Social and Emotional Assessment (ITSEA) Imitation/Play and Social Relatedness scores.

Results: The mean Total STAT score for this sample was 3.0 (sd = 0.54). Total STAT scores correlated significantly with RJA (r = -.35, p = .016) and IBR (r = -.39, p = .006) scores from the ESCS; but not with IJA score (r = .03, p = .909). Total STAT scores also correlated with the number of toys used in the DPA (r = -.31, p = .032). In addition, the Total STAT score correlated with the Imitation and Imitation/Play subscales of the PIA-CV (r = -.39, p = .006) and the ITSEA (r = -.35, p = .015), but not with the Nonverbal Communication (r = -.05, p = .723) or Social Relatedness (r = .09, p = .525) scales.

Conclusions: Correlations between the Total STAT score and several observational and parent report measures of communication, play, and imitation provide initial evidence for the use of the STAT as a continuous measure of nonverbal social-communicative behaviors in this unique sample of very young children with autism symptomatology.  Replication with other samples and examination of its sensitivity to change will represent important next steps in exploring its utility as a continuous measure.