Validating design indicators for an AI-assisted teaching strategies training module for university EFL lecturers using the Fuzzy Delphi Method

Abstract

Artificial intelligence is increasingly used in language education, yet classroom adoption in higher education often remains tool-focused. Lecturers, therefore, receive limited support on how to translate artificial intelligence into coherent and transferable teaching strategies. This study aimed to propose an artificial intelligence–assisted teaching strategies training module for university English as a foreign language lecturers and to validate design indicators to guide the next development stage. Forty-eight draft indicators were organised into five dimensions: objectives, content, unit structure, materials and resources, and assessment methods. Indicator validation used the fuzzy Delphi method with a panel of 15 experts in language pedagogy, teacher education, educational technology, and assessment. Experts rated each indicator, and consensus was evaluated using a predefined distance threshold, percentage agreement, and a fuzzy score. Overall agreement was strong. All indicators for objectives, unit structure, and materials and resources were retained (17/17). Most content indicators were retained (23/24). Assessment indicators were retained more selectively (4/7), reflecting higher expectations for feasible, classroom-ready evidence of strategy use. These findings specify essential design requirements for a strategy-oriented training module and support evidence-informed decisions in subsequent development and later pilot evaluation in authentic teaching settings. The indicator set also provides clear specifications for developing unit activities, resources, and assessment artifacts that can be reviewed and refined through iterative trials.