Using Multiple Data Streams in Executive Function Training Games to Optimize Outcomes for Neurodiverse Populations.

To be optimally effective, digital technologies should be adaptive to specific learners’ needs. Two examples are presented of data-informed approaches to developing digital games that support the development of executive functions (EF) in neurodiverse populations. The first is an experiment with younger and older adolescents that compared two versions of a video game designed to train the EF skill of inhibition. Based on developmental neurocognitive differences, one version focused on the speed of learners’ responses, while the other focused on the accuracy of responses. Results indicated that, as hypothesized, younger adolescents benefited more from the focus on speed, while the older adolescents benefited more from the focus on accuracy. In the second example, ongoing work on adapting an EF game designed to train the EF skill of shifting for high-functioning adolescents with Autism Spectrum Disorder (ASD) is presented. A detailed analysis of game log data, specifically data on speed and accuracy of responses in the game, revealed that although accuracy was near ceiling, there was greater variability in speed of responses. This suggests that for high-functioning adolescents with ASD, a version of the EF game that focuses on speed of response would be most beneficial. Next steps for the project are discussed, as are broader implications for data-driven approaches to designing adaptive digital tools for learning.