Data and Reproducibility Materials for "Adaptive" in Name Only? A Systematic Review and Multilevel Meta-Analysis of AI-Based English Language Learning Systems for K-12 Learners

Abstract

This repository contains the research data, analysis code, coding records, search documentation, and reproducibility materials supporting the systematic review and multilevel meta-analysis entitled “Adaptive” in Name Only? A Systematic Review and Multilevel Meta-Analysis of AI-Based English Language Learning Systems for K-12 Learners. The final primary meta-analytic dataset comprises 26 independent study clusters and 74 language-learning effect sizes. The repository includes the final effect-size dataset, study-selection decisions, database search strategies, human adaptivity and risk-of-bias coding, effect-size construction and consistency-audit documentation, scale-resolution data, R analysis code, and principal model outputs and diagnostic figures. The primary analysis uses a three-level random-effects meta-analysis with cluster-robust variance estimation (CR2) and Satterthwaite small-sample correction. Effect sizes are expressed as Hedges’ g. Materials required to reproduce the reported primary model, sensitivity analyses, heterogeneity estimates, publication-bias analyses, and descriptive adaptivity analyses are included. Contents meta_analysis_dataset_FINAL.xlsx — final analysis dataset and study-level metadata human_coding_FINAL.xlsx — final human-coded adaptivity and risk-of-bias data fulltext_decisions_FINAL.csv — final full-text eligibility decisions search_strategy_FINAL.md — database search strategies and search documentation protocol_and_coding_rubric_FINAL.md — review protocol and adaptivity coding framework effect_size_audit_FINAL.md — effect-size construction and consistency-audit summary scale_resolution_extraction_FINAL.xlsx — exploratory outcome-scale-resolution data analysis_FINAL.R — analysis code meta_analysis_results_FINAL.xlsx — principal numerical analysis outputs diagnostic figures and repository documentation The files are provided to facilitate transparency, verification, and reproducibility of the reported systematic review and meta-analysis.