An entropic measure for the teaching–learning process

Abstract

Few phenomena are as complex as the teaching–learning (TL) process. The instruction efficiency and information on the state of knowledge of the student group are some key variables in this process. Guided by Shannon’s definition of information we propose an entropy based performance index ( S p ) for monitoring the teaching–learning process. Our index is based on item response curves (IRCs) which have been recently employed in physics education research. Our proposed index is an explicit function of the ability θ . A preliminary survey indicates that S p is low (high) for high (low) ability student groups. We propose a simple model to explain this. We have also carried out a number of controlled studies to study the dependence of S p on student ability, peer instruction and collaborative learning. Our studies indicate that S p plays a role analogous to entropy in statistical mechanics, with student ability being akin to inverse temperature, peer instruction to an ordering (magnetic) field and collaborative learning to interacting components.