Artificial intelligence and teacher competence: a scoping review of assessment, analytics, and professional development
TL;DR
The review found frequent use of supervised machine learning, regularized regression, EFA, CFA, SEM, and learning analytics, but limited reported use of explainable AI, subgroup fairness analysis, multimodal validation, and longitudinal designs.
Abstract
This scoping review maps how artificial intelligence (AI) is being connected to teacher competence in recent research. The review was based on 33 peer-reviewed articles published in 2022–2026 and identified through a bounded Web of Science search. Its purpose was not to evaluate intervention effectiveness, but to describe the extent, range, and nature of the available evidence on AI, machine learning (ML), and learning analytics (LA) in teacher assessment, modeling, and professional development within this indexed corpus. The mapped literature suggests two broad lines of work. One uses AI, ML, LA, and computational psychometrics to assess teaching practice, model teacher development, or measure AI-TPACK-related competence. The other treats AI as part of what teachers themselves need to know and do. Instruments represented in the corpus, such as TAICS, T-GAIC, AI-SRLS, AI-TPACK, and RAIS, broaden the concept of competence to include AI literacy, self-efficacy, ethical reasoning, readiness, and teacher–AI co-teaching. The review found frequent use of supervised machine learning, regularized regression, EFA, CFA, SEM, and learning analytics, but limited reported use of explainable AI, subgroup fairness analysis, multimodal validation, and longitudinal designs. Quality appraisal indicated stronger support for measurement-structure claims than for causal claims about professional-development effectiveness or high-stakes AI deployment.