Modeling Student Problem Solving for Improving Project-Based Learning

Abstract

Project-Based Learning courses are widespread in engineering curriculums worldwide. As such, understanding how students problem-solve in these settings benefits curriculum designers, educators, and future learners. The objective of this research was to employ a human factors approach of modeling and error analysis to develop and apply a cognitive model to illustrate how students make decisions and problem-solve in a Project-Based Learning course. In addition, the Systematic Human Error Reduction and Prediction (SHERPA) was applied to identify errors in the process illustrated through the model. Data was collected from direct observations of 84 students in the classroom over two terms, along with qualitative reflection surveys and individual interviews to understand the project's impact on their problem-solving skills during their studies. The final model can be extended to other PBL courses to allow the targeted design of course materials to fit the student problem-solving processes, improve the project, and thereby improve learning outcomes. One main finding of this study was that students consistently failed to predict consequences, and predominately relied on each other as well as course PowerPoints to solve any problems that arose, illustrating the value in developing processes for improved materials.