Adaptive Perceptual Learning in Electrocardiography: The Synergy of Passive and Active Classification

Recent research suggests that combining adaptive learning algorithms with perceptual learning (PL) methods can accelerate perceptual classification learning in complex domains (e.g., Mettler & Kellman, 2014). We hypothesized that passive presentation of category exemplars might act synergistically with active adaptive learning to further enhance PL. Passive presentation and active adaptive methods were applied to PL and transfer in a complex real-world domain. Undergraduates learned to interpret real electrocardiogram (ECG) tracings by either: (1) making active classifications and receiving feedback, (2) studying passive presentations of correct classifications, or (3) learning with a combination of initial passive presentations followed by active classification. All conditions showed strong transfer to novel ECGs at posttest and after a one-week delay. Most notably, the combined passive-active condition proved the most effective, efficient, and enjoyable. These results help illuminate the processes by which PL advances and have direct implications for perceptual and adaptive learning technology.