Strategies for Generative AI-Assisted Question-Chain Design in Senior High School English Reading-for-Thinking Lessons

The Reading and Thinking section of senior high school English textbooks is where the curriculum’s activity-based approach is expected to be realized most fully, yet its potential is often constrained by a single ordinary decision: what to ask. This paper proposes a three-layer framework and six teaching strategies for designing question chains in such lessons with generative artificial intelligence (GenAI). The framework aligns question functions with the curriculum’s progression from learning and understanding, through applying and practicing, to transferring and creating, and specifies what GenAI may contribute at each layer and what must remain the teacher’s responsibility. The six strategies cover the design cycle: prompting by cognitive function rather than by content, treating AI output as candidate questions rather than as a script, anchoring questions in textual evidence, making dependencies between questions explicit, anticipating student responses and misreadings, and reserving the culminating question for the teacher. A complete chain designed for one textbook text illustrates the strategies. The paper argues that GenAI expands the supply of candidate questions without displacing the teacher’s interpretive authority: the value of AI-assisted questioning lies less in any single generated question than in the architecture of the chain.Subject AreasEducation, Linguistics