AI in mathematics education

Important results from the project

The pilot project has aimed to both strengthen students ´knowledge of mathematics and increase partners´ knowledge about machine learning AI in classroom contexts. An important goal has been to see whether AI-based teaching materials, to a much greater extent than is possible for an individual teacher, can take into account students´ knowledge gaps and personalize the learning. Purpose and goals of the project have been met even though expected effects have not been achieved to the extent expected.

Expected long term effects

There are indications that certain groups of students have improved in mathematics. On the other hand, survey follow-ups with teachers and classroom observations have not been able to show that the teaching aid has adapted in ways clear for teachers and students. Hence, the project has raised questions about technology´s lack of transparency, ethical dilemmas that can arise when errors are "instilled" in the algorithms. Overall, the project has contributed to both increased knowledge for the project owner and all parties involved, but also improved the quality of the digital AI tool.

Approach and implementation

The project has been carried out in four phases, where the first was to adapt the AI tool to what the intervention would test and invite teachers to participate. Between October 2020 and February 2021, 2 6-week interventions were conducted in 22 classes from years 2, 5, 8. After the interventions we had questionnaires for the teachers. We conducted classroom observations and interviews with project participants that will culminate in an academic article. The students´ results will be compiled during the autumn of 2021 and have and will be disseminated internally and externally.