Computer Based Assessments of Executive Functions in Preschoolers with and without ASD: The Relations to Parent Ratings of Social and Behavioural Functioning

Background:  Executive Functions (EF) refer to higher cognitive processes involved in the conscious control of thought and action.  Compared to mental age (MA) matched peers, children with ASD, even those with average IQ and verbal ability, exhibit EF difficulties with inhibiting socially inappropriate actions, shifting attention flexibly, and generating goal-directed behaviour.  These early EF deficits predict later perspective taking problems (Pellicano, 2010) and behaviour and emotion dysregulation (Zingerevich & LaVesser, 2008).  Although deficits in EFs are characteristic of several developmental disorders, we hypothesize that in children with ASD, performance on EF tasks may be specifically related to aspects of social adaptation.

Objectives:  To develop EF profiles using computer-based tasks among MA matched preschoolers with and without ASD, and explore how they relate to parent ratings of social/adaptive and behavioural functioning.

Methods:  Forty preschoolers (ASD=20; TD=20), aged 36-83 months, will participate.  Participants with ASD have a clinical diagnosis of ASD, and exceed cut-off scores on both the ADOS and Autism Spectrum Rating Scales-short form (ASRS).  Participants in the TD group fall below ASRS cut-off.  The Stanford-Binet Intelligence Scales will assess IQ.  The EF battery consists of computer-based measures of inhibition (Boy-girl Stroop, Go/No-Go, and Preschool Continuous Performance Test) and working memory (Self-ordered Search; Kerns & McInerney, 2007).  Parents will complete the Vineland Adaptive Behavior Scales (VABS-II), Behavior Assessment System for Children (BASC-2), and ASRS-short form.

Results:  Preliminary results (TD=12) revealed positive correlations between the BASC-2 and VABS-II externalizing scales, as well as between the social skills subscale and socialization scale, respectively.  Moreover, we found a reciprocal negative correlation among the social and externalizing scales between the two instruments.  Self-Ordered Search (SOS) was associated with the externalizing and social indices of both measures, and with the BASC-2 behavioural symptoms scale.  The Continuous Performance Test was negatively correlated with VABS-II socialization and positively correlated with the externalizing scale.  Similar correlations are expected in the autism sample.  Based on pilot data with a 3-year-old with autism, we predict that SOS performance will not differ significantly across groups, likely due to the superior visual-spatial search abilities often observed in this population.  These children may correctly and efficiently select targets even though they must keep in mind a series of previous search choices, both correct and incorrect.  However, we anticipate that children with ASD will perform significantly less well on inhibition tasks compared to the TD sample.

Conclusions:  This study represents a significant advancement with regard to the measurement approach and developmental period assessed.  In children with ASD, computer based, as compared to manual tasks of EF, may provide a more valid assessment of EFs as more precise information on reaction time is provided, the computer tasks are more intrinsically motivating, and socially-mediated aspects of task administration are reduced.  The preschool years are an ideal developmental period to assess EFs, as this is a time in which cognitive control demands significantly increase, particularly in social contexts.