Other
Artificial intelligence, student engagement, well-being and self-regulated learning in school education: Systematic review dataset and supplementary analytical materials
Beatrice Doria, Grazia De Angelis, Elena Dell’Aquila, Agostino Carbone, Angelo Rega et al. ·
Summary
Written by AI, not yet AICS's view · claude-sonnet-5-5This record is the dataset and supplementary material for a systematic review of how artificial intelligence tools used in schools relate to student engagement, well-being and self-regulated learning. The review covers early childhood through upper secondary, vocational and special education, and excludes higher education. Four databases were searched, and 84 empirical studies were included after screening 3,624 records and assessing 130 full-text reports. The record provides the codebooks, screening decisions, inter-rater agreement data, lists of included studies and co-occurrence analyses of outcomes and mechanisms. It does not itself report the review's findings in the text provided.
Key points
- The review looks at AI applications in school education and their link to engagement, psychological and socio-emotional well-being, and self-regulated learning.
- Databases searched were Scopus, Web of Science, ERIC and APA PsycInfo.
- Filters were peer-reviewed journal articles, full text available, English or Italian, published 2014 or later.
- 4,304 records were retained after filters; 3,624 were screened after duplicate removal, 130 were assessed in full text, and 84 empirical studies were included.
- Annex A holds the PRISMA flow data, screening decisions and inter-rater agreement data; Annex B holds the included-study list, analytical codebook, ATLAS.ti matrices and outcome × mechanism co-occurrence analyses.
Limitations
The text is a description of a dataset and supplementary files, so it gives no actual findings on whether AI tools help or harm children's engagement or well-being. The review is limited to English or Italian peer-reviewed articles from 2014 onward, and the Italy country tag is only a hint, as the review's geographic coverage is not stated.
Abstract
This record contains the dataset and supplementary materials of a systematic literature review examining how artificial intelligence (AI) applications used in school education relate to students' engagement, psychological and socio-emotional well-being, and self-regulated learning (SRL), the pedagogical and psychological mechanisms through which these effects are reported to occur, and the contextual, learner-related, teacher-related and technological conditions that shape them. The review covers early childhood, primary, lower and upper secondary, vocational and special education settings; higher education was excluded by design. Records were retrieved from Scopus, Web of Science, ERIC and APA PsycInfo (EBSCO) using a search string combining AI-related terms (e.g., artificial intelligence, machine learning, learning analytics, intelligent tutoring, adaptive learning, chatbots, conversational agents, generative AI, large language models), school-education and population terms, and well-being, emotional, motivational, engagement and self-regulation terms. After the application of formal filters (peer-reviewed journal articles, full-text availability, English or Italian language, publication year ≥ 2014), 4,304 records were retained; following duplicate removal, 3,624 records were screened by title and abstract, 130 full-text reports were assessed for eligibility, and 84 empirical studies were included. Screening was conducted independently by multiple reviewers, and inter-rater agreement data for both screening stages are provided. Annex A (Systematic Dataset) contains the extraction codebook, the PRISMA flow data with reasons for full-text exclusion, the bibliographic records exported from each database, the screening decisions at the first (title/abstract) and second (full-text) stages, the inter-rater agreement data, and the list of included articles. Annex B (Supplementary Material) contains the list of included studies (Table S1), the analytical codebook with operational definitions and inclusion/boundary rules (Table S2), the code-document matrices exported from ATLAS.ti for each research question and analytical domain, the characteristics of the included studies (geographical area, research methodology, AI system, educational context), and the outcome × mechanism co-occurrence analyses (absolute co-occurrence, conditional percentages, Jaccard similarity, phi coefficients, interpretive summary and heatmap data).