Tapping into the Pedagogical Potential of infinigoChatIC: Evidence from iWrite Scoring and Comments and Lu & Ai’s Linguistic Complexity Analyzer

Abstract

ChatGPT has drawn rapid interest among linguists and language teachers, with academics and industries rushing to tap into its potential. Yet the pedagogical potential of ChatGPT for L2 English writing remains to be further explored academically. This study aims to examine the effects of ChatGPT, or one of its Chinese versions called infinigoChatIC, in improving the quality of L2 writing and in facilitating the development of L2 proficiency among some postgraduate students learning an English reading and writing course in a prestigious university in Southwest China. Specifically, the main research questions focus on how infinigoChatIC-polished L2 English writings perform in comparison to L2 learner writings, and how the improvement of English writing is reflected in the changes of lexical and syntactic complexity, as evidenced by measured data and its pedagogical significance. A set of 24 English writings are collected as part of regular homework assignments, then this set of English writing is polished by the teacher researcher in the infinigoChatIC dialogue box with the prompt “please correct grammatical errors in the following English text so that it is grammatically accurate.” These two sets of writings are scored in iWrite automatic scoring system and then analyzed in Lu & Ai’s web-interfaced linguistic complexity analyzer. The study finds that infinigoChatIC can effectively improve the quality of L2 English writing, with lexical complexity vastly improved and yet surprisingly overall syntactic complexity slightly reduced. Pedagogical implications are discussed.