The Educational Puzzle
Clinical reasoning is often described as the “holy grail” of medical education. We know it matters. We know students struggle with it. Yet how best to teach it, especially to beginners, remains debated.
Our study, CREST (Cultivating Reasoning Through Example-Based or Self-Explanation-Based Teaching), set out to compare two theory-driven approaches.
The first, Self-Explanation (SE), encourages students to explain aloud why clinical signs occur, linking basic science to patient presentation. The second, Example-Based Teaching (EBT), shows students fully worked examples of expert reasoning, with tutors thinking aloud to model how diagnoses are made. While both methods are supported by research, they had rarely been directly compared in early ophthalmology training.
Ophthalmology provided an ideal testing ground. Retinal diseases such as diabetic retinopathy, age-related macular degeneration, and retinal vein occlusion often present with overlapping features, making diagnostic reasoning essential.
We randomly assigned novice medical students to learn through either SE or EBT. We assessed their reasoning using structured case-based questions and measured cognitive load to understand how mentally demanding each approach was.
Ten days after teaching, students in the EBT group performed better and reported higher germane cognitive load, the productive mental effort involved in building diagnostic frameworks. By forty days, the difference disappeared, raising new questions about consolidation and long-term learning.
CREST does not claim a universal winner. Instead, it highlights the importance of timing and learner level. For novices, explicitly modelling expert thinking may provide a clearer entry point into complex reasoning.
Ultimately, this research is about patient safety. Better reasoning leads to better diagnoses and better outcomes.