Learning Schemas from Explanations in Practical Electronics

Training materials in practical electronics appear to follow a building blocks approach in which common simple circuits are presented and then combined into more complex circuits. Each circuit is presented in the form of a circuit diagram and an explanation of how the circuit works in terms of a causal chain of events. Such materials suggest that learning electronics consists of learning schemas for the building block circuits; complex circuits can then be understood as combinations of these simpler schematic circuits. The process of learning appears to be based on extracting schemas from the explanations. This report presents human experimental results based on earlier artificial intelligence work in this project. Engineering students learned building block circuits and then learned complex circuits; the time required to understand the explanations and answer questions about the circuit behavior were compared to an AI system that learned from explanations and a model of question-answering. Generally, learning the schematic building block circuits facilitated performance, and the AI system and question-answering model could predict the amount of facilitation. However, the benefit of learning circuit schemas under these conditions was surprisingly mild.