Application possibilities of generative artificial intelligence in mathematical error analysis

Abstract

Generative artificial intelligence (GenAI) has the potential to support teachers in routine tasks. This study investigates the use of GenAI in diagnostic activities using 200 simulated student errors in written arithmetic operations. The error analyses were carried out once with a basic prompt and optimised in a second phase using prompt engineering. Overall, the first positive results are evident, but the results make it clear that GenAI is not yet fully developed for reliable diagnostic support in the classroom without specific training.