Design chatbots for differentiated engagement across proficiency levels, beyond simple error correction¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The authors conclude that chatbot design should respond to proficiency-based differences in interaction. Because high-proficiency learners sustain dialogue and negotiation while low-proficiency learners cycle through corrective feedback, chatbots "must engage with varying EFL proficiencies in a way that transcends simple error correction and stimulates the deep cognitive processing necessary for language acquisition". Future work should investigate why low-proficiency learners rely on the check_button so systems can encourage them beyond the corrective loop.

Design Implications¶

Context¶

Requirements¶

• Behavioural log analysis (e.g. TNA) to identify proficiency-specific interaction patterns before designing differentiation

Constraints¶

• The study is cross-sectional and cannot establish causality; the finding comes from one cultural and educational context with a specific chatbot and LLM model

Target Learners¶

• EFL learners across proficiency levels, especially low-proficiency secondary students

Target Learning Goals¶

• Deep cognitive processing and negotiation of meaning in L2 writing, not only surface error correction

Affordances¶

• Transition Network Analysis Framework

Related Strategies¶

Examples¶

-

Key Sources¶

• Steve Woollaston, Brendan Flanagan, Yuko Toyokawa and Hiroaki Ogata. (2026). Penny: Transition Network Analysis of Learner-Chatbot Interactions in Scaffolded EFL Writing. Preprint, accepted at the 1st Workshop on Transition Network Analysis (TNA) at LAK '26. https://arxiv.org/abs/2607.14575