Diversity of differentiated (functionally appropriate) object play is thought to index greater object knowledge in young children exhibiting signs of autism. It is often defined as the number of unique play actions observed during an assessment (Yoder, 2006), and has been shown to positively predict later language development in children with autism (Sigman & Ruskin, 1999; Yoder, 2006). However, little is known about assessing play behavior in very young, non-verbal children (less than 24 months) with autism symptoms.
Objectives:
This study examines the validity of an observational measure of object knowledge in toddlers exhibiting early symptoms of autism.
Methods:
Participants from the initial assessment of A Multi-Site Clinical Randomized Trial of the Hanen More than Words Intervention included 63 toddlers (mean CA = 21.2 months, range = 15.5 – 25.0 months). They met a predetermined cutoff on the STAT and had a clinical presentation consistent with ASD. A nomological network approach (Cronbach & Meehl, 1955) and a divergent validation approach were utilized to assess the construct validity of a proposed measure. Diversity of play was defined as the number of differentiated play actions by the child during a 7-minute portion of the Developmental Play Assessment. The constructs predicted to be associated with diversity of play actions were (a) attention, (b) mastery motivation, (c) play and imitation, and (d) object-oriented turn-taking. The four variables used to represent these constructs were a) the Attention Subscale of the Infant/Toddler Social and Emotion Assessment (ITSEA), b) the Mastery Motivation Subscale of the ITSEA, c) the sum of the number of items failed on the Screening Tool for Autism in Two-Year-Olds (STAT) Play and Imitation domains, and d) the sum of the number of action turns and give turns by the child during an experimental measure of object turn-taking (Yoder & Stone, 2006). For the divergent validation approach, we examined these same correlates with frequency of initiating joint attention (IJA) in the Early Social Communication Scale (ESCS).
Results:
As predicted, the number of differentiated play actions was correlated with the ITSEA Mastery Motivation subscale, the number of items failed on the STAT Play and Imitation domains, and the sum of the action and give turns (r = .31, -.48, and .46 respectively). The number of differentiated play actions was not correlated with the ITSEA Attention subscale. To provide further evidence of validity, relationships between the 4 comparison variables and a positive, but theoretically unrelated variable, the IJA total score from the ESCS, were tested. IJA was not significantly correlated with attention, mastery motivation, play/imitation, or turn-taking.
Conclusions:
These results support the hypothesis that the number of differentiated play actions by a child may be a valid measure of object knowledge in toddlers showing early signs of autism. These findings are encouraging, given that there are currently no measures of early object knowledge validated for use with very young, non-verbal children. This study provides support for further examination and validation of the diversity of play measure with this unique population.