Did AI-supported inquiry improve students' climate-change reasoning?
Evidence level: Level 2: Correlational. This study compared groups with baseline data, which would normally reach Level 3. We lowered it one level after reading the full paper: whole classes were assigned to each condition, and the analysis did not adjust for that. How we grade
Source: Can AI-Assisted Inquiry Enhance Students' Decision-Making Skills in Socio-Scientific Issues? A Three-Group Experimental Study on Climate Change, arXiv, 2026-07-26. The arXiv record links a journal version (DOI 10.62225/2583049X.2026.6.3.6516), and the paper's pages carry the header International Journal of Advanced Multidisciplinary Research and Studies. We have not confirmed that journal's review process.
The study in plain language
Researchers compared three groups of secondary students (270 in all, 90 per group, grades 8 to 11) learning to make decisions about climate change: inquiry with a question-oriented AI assistant, the same inquiry without AI, and traditional teacher-led instruction. All three groups started at similar, mostly low levels and all improved. The AI-assisted group improved the most, with large effects compared with traditional instruction and moderate-to-large effects compared with inquiry alone. The study did not compare different AI designs, so it cannot show that a questioning AI works better than one that gives answers. The post-test came soon after a short intervention, so it does not show whether the gains lasted (full paper).
What the abstract says
Climate change is a socio-scientific issue: it rests on science but cannot be settled by science, because any serious response forces people to weigh costs, values, and competing interests under uncertainty. Helping students make such decisions well is a central aim of science education, and the arrival of generative artificial intelligence raises a sharp question: does a conversational AI partner deepen students' reasoning, or simply do the thinking for them? This study tested whether AI-assisted inquiry improves secondary students' decision-making about climate change. Using a pretest-posttest design with three groups (AI-assisted inquiry, inquiry without AI, and traditional instruction; 270 students, 90 per group), reasoning was assessed across seven decision-making steps, from defining the problem to monitoring with adaptive management, using a four-level analytic rubric scored through content analysis with high inter-coder agreement. All three groups began at comparable, mostly low levels and all improved, but the gains differed sharply. The AI-assisted group improved most, ahead of inquiry-only and of traditional instruction. Between-group effect sizes on gains were large for AI-assisted versus traditional instruction and moderate-to-large for AI-assisted versus inquiry-only, with the clearest advantages on stakeholder engagement, alternatives, implementation, and monitoring. Within the AI group, the number of times students checked the AI's claims against the sources predicted their gains, and no student was flagged for over-reliance. The findings suggest that AI helps most when it is designed to question rather than to answer, and that the inquiry it is embedded in carries much of the benefit.
Quoted from the abstract at arXiv.
What it does and doesn't tell us
What it tells us
- Who took part: 270 secondary students in grades 8 to 11 (90 per group).
- What was measured: the quality of students' written decision-making reasoning about climate change, scored with a four-level rubric the researchers defined for each of the seven decision-making steps.
- The paper reports that experts reviewed the instruments for content validity and that agreement between coders was substantial to near-perfect.
What it doesn't tell us
- The design was quasi-experimental: intact classes, not individual students, were assigned to conditions. Students in the same class are not fully independent, and the reported tests did not adjust for this, so they may look more precise than they are. The authors note this in their limitations. The paper does not report how many classes were in each condition.
- A delayed post-test was not reported, so we do not know if the advantage lasted.
- Written worksheets may favor students who write fluently, as the authors also note.
- It does not tell us whether the result holds for learners, subjects or settings the study did not include.
Read it yourself
Read the full study at arXiv. Evidence levels describe study design and who took part. They are not endorsements or official ratings.