Unravelling the Connection: Online Learning Behaviors, Task Difficulty, and Preservice Teachers' Performance on a Computational Thinking

The computational thinking of preservice teachers has been a subject of extensive investigation, but with mixed outcomes regarding the effectiveness of computational thinking training. The goal of this project was: (1) to examine how online learning behaviors and perceptions of difficulty level would predict preservice teachers’ performance on a computational thinking challenge and (2) to determine how preservice teachers’ attitudes toward computational thinking would vary based on their online behaviors. This research project introduced an online training playground and then used Decision Tree, a supervised machine learning algorithm, to categorize preservice teachers using both log data and survey responses. Total time spent on the playground, total time spent on solving the challenge, and individual perceptions of difficulty level after having completed the challenge emerged as the three most significant predictors of performance. Participants varied across their computational thinking based on the time they spent on the playground and on the actual challenge. The importance of using log data such as online learning behaviors to identify areas of improvement in teacher-education are discussed.