1 nota · última: 11 Nov 2019
Research from the University of Arizona argues that learning is maximized when failure happens about 15% of the time, with performance around 85% accuracy in simple two-choice tasks.