Effects of Statistical Training on Children with Autism Spectrum Disorder
Sulynn Miao · eScholarship (California Digital Library) · 2016
SURF Conference Proceedings 2016 Session 2A Author: Sulynn Miao Faculty Advisor: Fei Xu Effects of Statistical Training on Children with Autism Spectrum Disorder Introduction: This research looks at Autism Spectrum Disorder, or ASD, to see if a brief statistical training influences probabilistic reasoning skills in children with ASD. Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that affects many different parts of an individual’s functioning. It is characterized by impairments in social functioning and language, such as communication and interaction skills. Individuals with ASD struggle with generalization, which is the ability to extract a general principle or rule based on small amounts of data. This skill is incredibly important when it comes to learning; in fact, much of how we learn about the world involves generalizing from the information around us. In order to make such generalizations it’s necessary to have an understanding of intuitive statistics, which allows human learners to make predictions based on the data given. Previous Research: Sim & Xu (2016) explored this understanding of intuitive statistics, and found that that children with ASD show weaknesses in probabilistic reasoning as compared to typically developing (or TD) children. They conducted a study showing both TD children and children with ASD videos of objects bouncing around inside circles (Figure 1.1) of three yellow objects and one blue object. During the test trials, the circle would be blocked as well as the outcome, and the child would be shown two outcome cards (Figures 1.2 &1.3) and had to choose which one they think would happen; would the majority (yellow) object be the one that fell out, or the minority (blue) object? Figure 1.1 Sim & Xu (2016) Figure 1.2 Outcome 1 Figure 1.3 Outcome 2 The results of the study showed that children with ASD did show impairments in probabilistic reasoning, meaning they weren’t able to correctly infer that the majority objects were more likely to fall out. This could allude to something very important about the underlying mechanisms of ASD being rooted in the way children with ASD learn new information, which would have cascading consequences in all domains of learning throughout their lives. This study leaves the need for exploration of ways in which children with ASD could acquire such skills of probabilistic reasoning. Stanley & Lawson (2014) sought a similar goal, but with typically developing (TD) children. They found that an explicit training on statistical principles for TD children increased their ability to accurately use these skills. Goals: