Generating Competitive Distractors from Student Error Data

While students often prefer learning with multiple-choice questions (MCQs), these are frequently criticized for emphasizing recognition over active recall, which makes them less effective for long-term retention. However, well-designed competitive distractors, that are semantically or orthographically similar to the correct answer, may overcome this deficit by encouraging deeper cognitive processing. Traditional distractor generation is labor-intensive, requiring domain expertise and significant time investment. Recent advances in artificial intelligence offer automated solutions for generating semantically related distractors at scale, but not for generating distractors that target learners' misconceptions about the lexical representation of answers. In this study, we present a data-driven method for automatically generating plausible orthographically related distractors, based on a combination of common incorrect responses from learners doing open-answer retrieval practice, and rule-based generation of common misspellings. This flexible and scalable approach to generating distractors that align with genuine misconceptions, can make MCQs a more effective tool for vocabulary learning.