27940
A Mobile Game Platform for Studying Social Influences on Executive Function in ASD: Towards Accessible Remote Measurement
Mobile applications and games offer an accessible, often highly acceptable, method for assessing skills and developmental ability in special populations (e.g. young children), potentially facilitating data collection from the large sample sizes needed to parse and study clinically-relevant heterogeneity. Here, we present a mobile game that uses social and nonsocial stimuli to examine executive functioning skills, a known vulnerability for children with ASD.
Objectives:
1. Design a fun tablet-based game targeting areas of executive function.
2. Compare game performance and play patterns between diagnosis groups (ASD, typically developing (TD)) across stimuli (social, nonsocial).
3. Assess the predictive value of game performance to IQ and a validated measure of executive function, the Behavior Rating Inventory of Executive Function (BRIEF).
Methods:
Three games were created to assess: shifting, working memory, and inhibitory control. The shifting game showed four images (happy or angry faces, red or blue fractals) and required users to guess which image fit the unwritten rule. The working memory task required users to memorize one to four images and then choose the correct objects. The inhibitory control game required users to touch as many pictures as possible, but stop when a stop sign or angry face appeared. Each game had two social blocks and two nonsocial blocks that alternated (three counterbalanced orders). Participants were sixty-five children (ASD, N=33; TD, N=32) ages 2 to 17 years. Experimenters placed the tablet in front of participants seated at a table. Directions were presented on the screen and read aloud. Participants were allowed to engage with the tablet in any way they preferred.
Results:
A one-way ANOVA revealed, in the shifting game, participants with ASD made more mistakes than TD children in social (p=.015) and nonsocial conditions (p=.021). Children with ASD also touched more targets than TD children during the stop period of the inhibitory game in the nonsocial condition (p=.021), but not the social condition (p=.411). Inhibitory game performance was the only predictor of IQ (social, r=.345, p<.05; nonsocial, r=.456, p<.001). A one-way ANCOVA controlling for IQ and age revealed children with ASD were less accurate than TD children in the working memory game for social (p=.024) and nonsocial (p=.01) conditions. Partialling for IQ and age, significant relations were found between the BRIEF shifting subscale and shifting game duration (r=.467, p=.044) and correct guesses (social and nonsocial, r=.472, p=.035). Planning/organization subscale was correlated with nonsocial learning latency for the shifting game (r=.504, p=.028). In the inhibitory game, nonsocial accuracy was correlated with self-monitoring (r=-.421, p=.041) and emotional control (r=-.402, p=.05) subscales.
Conclusions:
In support of previous findings, children with ASD showed several EF deficits compared to their TD peers. In some areas, these deficits may interact with the sociality of the task. Mobile games, including the game presented here, have the potential to predict valuable clinical indicators such as IQ. While some facets of the BRIEF correlated with some game outcomes, more exploration is necessary to elucidate these relationships.