Use of Decision‐Based Learning to Develop Conditional Knowledge of Binary Phase Diagram Interpretation in Mechanical Engineering

ABSTRACT Instruction and educational research on teaching multicomponent phase diagrams (PDs) to engineering students often focuses on visualization tools instead of promoting conditional knowledge, an understanding of when information is useful for a specific task. Decision‐Based Learning's (DBL's) emphasis on scaffolded decision‐making and just‐in‐time learning may offer a more effective approach. This retrospective study examined whether DBL improves student performance during PD instruction compared with traditional methods and investigated how students and instructors perceive its benefits and shortcomings. A mixed‐mode approach was used. Exam performance and survey data were analyzed for five semesters of an introduction to materials science course in mechanical engineering. The exam performance of 520 students was analyzed using analysis of covariance with a covariate to account for prior knowledge. Surveys and interviews provided insight into student and instructor experiences using an explanatory sequential design. Quantitative results showed improved performance during the first semester DBL was implemented, but no consistent effects across the other three DBL‐using semesters were seen. Qualitative results showed students valued structured feedback and expert‐modeled thinking, though some reported technical difficulties. Instructors found classroom walkthroughs of the DBL tree allowed for clearer scaffolding, providing improved engagement and understanding compared with alternative activities. Our findings suggest DBL can improve students' ability to interpret PDs compared with alternative teaching methods when DBL is actively used during class. However, this retrospective, rather than randomized‐controlled, study limits our ability to isolate the effects of DBL on performance and experience.