Rethinking Education through “Mesoscopic” Models of Learning
A “mesoscopic” learning model provides a functional description of how the brain carries out a specific type of learning task. In this talk I will introduce my research on two such models: the multimedia learning model (MLM) and the deliberate practice model (DPM). By combining existing multimedia learning theory with grounded cognition, a novel theory in cognitive psychology, I constructed an improved version of MLM that is especially useful in designing instructional animations in physics. In a clinical experiment, animations designed according to this new model led to significantly better learning outcome than those designed by the existing model. In a separate study, we applied both the improved MLM and the DPM to the design of online “deliberate practice” activities that focus on developing specific problem solving skills in physics such as mapping between physics situation and math expressions. Students in our online course trained via “deliberate practice” activities develop individual skills faster than those trained via isomorphic multiple choice problems. However, whether “deliberate practice” activity develops overall problem solving ability better than traditional calculation problems remains an open question. In addition, I will also share my ideas on improving the quality of assessment by constructing a mesoscopic cognitive model of problem solving, which could potentially
Read More