An Investigation of the Effectiveness of AI-Generated Contextual Texts for Extracurricular Vocabulary Acquisition among Tenth-Grade Students

作者：Liu Yumei, Jin Zixi · 发表于：Frontiers in Educational Research · 年份：2025 · DOI：10.25236/fer.2025.081031

: This quasi-experimental study investigates the effectiveness of AI-generated contextual texts extracurricular resources for vocabulary acquisition among 16 Chinese tenth-grade students with documented lexical difficulties. Grounded in the New Curriculum Standards' emphasis on vocabulary learning and the comprehensible input hypothesis, the research compared a traditional word-list memorization approach (control group) with an AI-mediated contextual learning approach using ChatGPT 3.5-generated passages (treatment group). Over three treatment cycles, participants engaged in 20-minute after-class study sessions, followed by productive vocabulary post-tests and semi-structured interviews. Despite non-significant statistical differences in post-test scores (p > .05) attributable to limited sample size and treatment duration, convergent analysis of quantitative and qualitative data yielded pedagogically significant findings. AI-generated texts demonstrated capacity to: (1) scaffold lexical learning through thematically coherent, difficulty-calibrated contexts aligned with the "i+1" principle; (2) promote deeper semantic encoding via output-driven assessment tasks; and (3) function as schematic bridges integrating new vocabulary with existing knowledge frameworks. Qualitative feedback revealed that while the control group experienced rapid engagement decline and lexical attrition, the treatment group reported enhanced contextual inferencing, accelerated acquisition, and sustained...