Students’ Critical Thinking Disposition in Artificial Intelligence-Assisted Mathematics Learning
Year: 2026
Type: article
Study design: Observational Study
Abstract: Artificial Intelligence (AI) can be used by students to learn and solve math problems in seconds. However, the learning habits of Indonesian students tend to rarely involve the ability to think at a higher level. This study aims to analyze students’ critical thinking dispositions in AI-assisted learning based on factor structure and experience category mapping. This study involved students (N=99) at a junior high school in the city of Jambi, Indonesia, so the findings are exploratory and limited to the context of the school. Data were collected using a perception-scale questionnaire on AI-assisted math learning behavior, analyzed with Exploratory Factor Analysis (EFA) and Rasch analysis. The findings identified five factors as the basis of student response patterns associated with critical thinking dispositions in AI-assisted math learning—reflective monitoring, internalization of understanding, passive dependence on AI, critical evaluation, and cognitive load. The mapping results showed that most students rarely experienced cognitive overload or internal understanding. At the same time, only a small percentage routinely performed reflective monitoring and critical evaluation, accompanied by early indications of a tendency toward passive dependence on AI. However, these findings need to be interpreted conservatively given the relatively low reliability of these factors. Based on the difficulty level of item approval in Rasch’s analysis, student responses showed inconsistencies in several factors, which could indicate a risk of cognitive passivity; However, these findings reflect the overall difficulty level in item approval, not a direct classification of each student’s level of disposition. This is especially true of the passive dependency factor on AI, which has the lowest reliability among the five factors. Therefore, strengthening working memory in AI-assisted mathematics learning requires an emphasis on reflection, evaluation, and justification activities to support knowledge development through internalization.
Institutions: Universitas Jambi, De La Salle University
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: diamond
OpenAlex ID: W7225576225