Associations of Student Response Times with Course grades in Introductory Physics courses.

While accuracy in basic math skills has been shown to predict the performance of students in introductory Physics courses, the predictiveness of response times (RT) on such outcomes remains largely unexplored. By examining data from online pre and post test assignments in the course, we find that RT can explain additional variance in student grades above accuracy alone. We find that we can also meaningfully categorize students based on their pre scores and RTs as well as their evolution to post test scores and RTs, and these categories have significantly different mean course grades. With regards to this evolution, we hypothesize that a meaningful improvement in speed (without any harm to accuracy) is more likely to happen if the student first reaches some minimum level of accuracy, and we provide correlational evidence to support this hypothesis. Finally, in a first step to better determine the evolution of performance and any causal link between RT and course grades, we explore the extent to which RT can be improved by using STEM Fluency, an online mastery practice application for basic STEM skills.