Reviewed studies reported good algorithm accuracy and efficiency for AI in STEM, with performance varying by algorithm and granularity¶
Claim · All claims
Evidence · 1 study · reviewr2·q2
Currency · newest study 2022 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2025: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? From the technological perspective, 24 of 63 reviewed articles reported effects such as efficiency of technology and accuracy of algorithm, mainly reporting good efficiency and algorithm accuracy. → Weiqi Xu and Fan Ouyang 2022
Evidence¶
Weiqi Xu and Fan Ouyang 2022¶
Weiqi Xu and Fan Ouyang. (2022). The application of AI technologies in STEM education: a systematic review from 2011 to 2021. International Journal of STEM Education. https://doi.org/10.1186/s40594-022-00377-5
q2 · i? · review · r2
The review's synthesis of technological findings covers 24 articles. Examples include AdaBoost.M1 achieving the best performance with the highest accuracy among machine learning algorithms grading open-ended physics questions, ensemble methods (GLMNET) achieving high accuracy where logistic regression performed poorly, and an automated assessment system whose performance degraded when scoring models were built on suites of items rather than individual items.
"From the technological perspective, 24 of the 63 reviewed articles reported the technological effect and findings (e.g., efficiency of technology, accuracy of algo- rithm) when applying AI techniques in STEM educa- tion."
Discussion¶
Related Claims¶
- Most reviewed studies reported significantly positive AI effects on STEM learning performance, with two null results
- Zero-shot LLMs perform only marginally above random in five-class competency classification but exceed 70% accuracy on binary classification
- AI tools collected and organized a previously unavailable granularity of data, including individual student exchanges and classroom discussion data, to generate actionable recommendations
- Machine learning methods outperformed logistic regression in classification accuracy across reviewed studies — a narrower finding that bears on this claim
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
ai-stem-technological-effects-algorithm-accuracy |
title |
Reviewed studies reported good algorithm accuracy and efficiency for AI in STEM, with performance varying by algorithm and granularity |
description |
Reviewed studies reported good algorithm accuracy and efficiency for AI in STEM, with performance varying by algorithm and granularity |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2022 |
recheck_by |
2025 |
generated |
by process:wiki-ingest on 2026-10-16 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — weiqi-xu-and-fan-ouyang-2022 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.