Critical Reader: A Contextual Vocabulary Trainer for Second Language Acquisition

Abstract

Many English language learners struggle to build the vocabulary and metacognitive strategies needed to adapt in new linguistic environments. This project addressed that challenge with Critical Reader, a custom ChatGPT-based instructional tool that helps learners develop vocabulary acquisition and retention habits.

The instruction begins with an introductory and “language hack” video, followed by guided practice using the Pre-Think, Assume, Use (PAU) cycle—a structured conversational routine. Grounded in Cognitive Load and Flexibility Theories (CLT & CFT), the GPT dynamically adjusts prompts based on learner level by adding context clues that support comprehension.

Evaluation data from four usability tests and 16 learner performance tests indicated the instruction met its objectives. Learners showed improved grammar and greater accuracy in inferring word meaning, while instructors reported faster vocabulary retention. Attitudinal survey responses reflected high engagement and perceived effectiveness.

Future improvements include API-based longitudinal learner performance tracking, expanded error analysis for personalized instruction, and multimodal features to support speaking and listening.