Improving Programming Content Delivery in an Introductory Biomechanics Course Using a Blended Classroom Approach

Abstract

MATLAB programming projects are assigned in the Vanderbilt University Introductory Biomechanics course to strengthen coding skill and demonstrate application of theory to practice. In course evaluations, students consistently rate these projects as the most challenging and difficult to grasp, citing the inefficacy of the traditional (lecture-only) course in merging programming with biomechanics content. Moving material online is an intuitive solution to this computation-based instructional challenge. Literature has shown that blended classes (30-80% online) can improve academic performance compared to in-person or online classes alone. While very few studies have analyzed blended learning in biomedical engineering contexts, research in related fields indicates students in blended engineering courses have improved attendance, motivation, and collaboration. We hypothesized that restructuring to a blended course would improve coding confidence and competence over the traditional course. Two courses were compared: one traditional course and another with programming content moved to weekly online modules. A programming project was assigned after completion of the coding material in each class. Modules were created using a backwards design approach. The desired coding competencies were identified as: pseudocode, loops, matrix operations, and data visualization. Modules for each of these subjects contained review, practice, and reflection components. Review and practice materials were designed by the authors or gathered from online coding resources and students reflected on the material by writing short responses speculating how the technique could be used to solve biomedical engineering problems. Following the modules, a project was assigned to assess coding skill by solving a hypothetical exercise biomechanics problem (resolving moment and angle about the elbow during a bicep curl) and displaying the solution graphically. To begin assessing the efficacy of the blended class method, students were surveyed on module effectiveness and their coding confidence. While responses indicated a lukewarm reception to the blended course, students in the blended class reported significantly increased coding confidence based on pre-and post-project surveys. Furthermore, students in the blended course reported significantly higher confidence using loops, performing matrix operations, and visualizing data compared to traditional course students. To assess coding ability, projects were re-graded by a teaching assistant using a standardized rubric to assess general code quality. While overall scores were similar between classes, students in the blended course had improved code readability. Students in the blended course also tended to write shorter code, indicating a blended class may improve students' ability to write well documented, organized, and efficient code. The results of this study provide evidence that blended content delivery can improve student programming confidence and performance compared to a traditional introductory bi...