AI-based Learning Assistants: Enhancing Math Learning for Migrant Students in German Schools

Increasing migration leads the education landscape to face the challenge of integrating children into the school system, accounting for inherent disadvantages. On average, as nearly half of them face socio-economic disadvantages, a persistent performance disparity was observed compared to domestic students. Remarkably, even after mitigating the influence of socioeconomic status, immigrant students' scores remained inferior. From March to May 2023, the AI Education (AIEDN) research project investigated how an AI-based learning assistant can compensate for existing impediments by enabling a better understanding through video-based learning. For this study, 275 students were selected from two secondary schools (N=137) and two grammar schools (N=138) in Baden-Wurttemberg, Germany. The experiment tested the extent to which learners solve more tasks, build broader (transfer) knowledge, and retain it. Students were repeatedly tested on mathematical problems, where the control group did not have access to the AI assistant. In contrast, the other group used the AI assistant to solve the problems. The students' performances from both school groups were statistically tested using t-tests. Within the test group of grammar schools, a significant change in the results between the two test days existed (T(41)=-2,28; p<.01) when the AI tool was used. In contrast, the control group showed no significant change T(37)=-.42; p>.05; d=.07) in performance. However, the results were insignificant for the secondary school students, as the given tasks and video content were considered too demanding. Further research is needed to determine AIEDN's performance for other target groups.