AI evidence extraction

The application of AI technologies in STEM education: a systematic review from 2011 to 2021

Weiqi Xu, Fan Ouyang · International Journal of STEM Education · 2022

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

5/10

Relevance

1/4

Quality (LMQS)

Evidence

404

Citations

74.38

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1186/s40594-022-00377-5

Methodology & findings

Study design

Systematic review following PRISMA principles.

Main result

The systematic review identified six categories of AI applications in STEM education: "learning prediction (N = 18, percentage = 29%), intelligent tutoring system (N = 16, percentage = 25%), student behavior detection (N = 13, percentage = 21%), automation (N = 8, percentage = 13%), educational robots (N = 6, percentage = 9%), and others (N = 2, percentage = 3%)". Regarding educational effects, "most of them showed significantly positive influence of AI techniques on the improvement of students' learning performances (N = 20)", and "most students held positive attitudes towards the use of AI technology in STEM education, and AI technologies aroused their interest and motivation as well."

Research paradigm

Mixed-methods interpretivism grounded in General Systems Theory

Author conclusions

"The application of AI technology in STEM education is confronted with the challenge of integrating diverse AI techniques in the complex STEM educational system. Grounded upon a GST framework, this research reviewed the empirical AI-STEM studies from 2011 to 2021 and proposed educational, technological, and theoretical implications to apply AI techniques in STEM education. Overall, the potential of AI technology for enhancing STEM education is fertile ground to be further explored together with studies aimed at investigating the integration of technology and educational system."

Risk of bias

Database selection bias: Only 8 major databases searched; potential for missing articles in smaller or specialized databases; Search term bias: Keywords may have favored technology-focused rather than education-focused research; Publication bias: Only peer-reviewed journal articles and books included; gray literature excluded; Language bias: Likely English-language bias, though not explicitly stated; Screening process bias: While inter-rater reliability was 92% for title/abstract screening, only ~30% of articles independently reviewed by second rater; Study design bias: Only empirical studies included; theoretical or opinion papers excluded; Geographic bias: No mention of geographic distribution of reviewed studies; Publication bias (only peer-reviewed articles from major databases searched); Selection bias in database selection (only 8 major publishers searched); Technology bias (studies emphasizing technological aspects over educational context); Language bias (searches likely conducted in English); Time period limitation (2011-2021 only); Geographic bias potential (major English-language databases primarily); Selection bias: potential for missing relevant studies due to keyword limitations; Publication bias: searches limited to peer-reviewed articles in English-language databases; Technology-centric reporting bias: studies may emphasize technological aspects over educational context; Geographic bias: predominantly English-language databases may underrepresent non-English research

Limitations

"Although we searched the best-known scholar databases with the keywords relevant to AI-STEM, some biases might exist in the searching and screening process
Since AI-STEM is a highly technology-dependent field, some studies might only highlight the technology rather than the education context
Therefore, future studies can adjust the searching criteria to solve these problems." Additionally, "we used a GST framework to examine the multiple elements in the complex AI-STEM system, but we did not investigate the mutual relationships between elements
Therefore, the complex relationships between different elements (e.g., instructor-learner, learner-learner relationship) in AI-STEM system need to be further explored."

Open questions raised

Need to investigate mutual relationships between system elements (instructor-learner, learner-learner) in AI-STEM
Expansion of AI applications to mathematics and engineering learning contents; currently concentrated in technology and science
Limited AI applications across educational levels beyond higher education; few applications in kindergarten and early childhood
Need for AI-empowered STEM learning environments combining advanced mediums (E-books, AR/VR) with traditional contexts
Future meta-analysis needed to report effect sizes and gain deeper understanding of AI-STEM integration effects
Need for adjustment of searching criteria to capture studies that balance technology and educational context emphasis

Extracted from: pdfAgreement 83%

Explore related topics

AI Literacy Training Doctoral Training