Abstract

Automated theorem proving based on proof planning is a new and promising paradigm in the field of automated deduction. The idea is to use methods and heuristics as they are used by human mathematicians and encode this knowledge into so-called methods. Naturally, the question arises whether these methods can be beneficially used in learning mathematics too. This paper investigates and compares the effect of different instruction materials (textbook-based, example-based, and method-based) on problem solving performance. The results indicate that the performance for the method-based instruction derived from automated proof planning in the WMEGA system is superior to that of the other instructions that were derived from a textbook and an example-based classroom lesson. These results provide a first support for introducing proof planning based on methodological knowledge into the school curriculum for mathematics.

Citations