An Active Learning Approach to Teach Advanced Multi-predictor Modeling Concepts to Clinicians

Abstract

Clinicians have characteristics - high scientific maturity, low tolerance for symbol manipulation and programming, limited time outside of class - that limit the effectiveness of traditional methods for teaching multi-predictor modeling. We describe an active-learning-based approach that shows particular promise for accommodating these characteristics.

• Active learning
• Statistics education
• Graduate education
• Deconstructing the disciplines
• Multi-predictor modeling
• Non-statisticians