A study into the use of Extended Concept Maps and Speech Recognition to improve lower-order language skills and higher-order thinking skills for language learners in Japan

Research Project

Project/Area Number	22K00760
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Review Section	Basic Section 02100:Foreign language education-related
Research Institution	Kanazawa University
Principal Investigator	Gary Ross 金沢大学, 薬学系, 准教授 (10708142)
Co-Investigator(Kenkyū-buntansha)	ヘネベリースティーヴン島根県立大学, 国際関係学部, 教授 (30405477) ヨークジェームズ明治大学, 政治経済学部, 専任講師 (90774498)
Project Period (FY)	2022-04-01 – 2026-03-31
Project Status	Granted (Fiscal Year 2022)
Budget Amount *help	¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000) Fiscal Year 2025: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000) Fiscal Year 2024: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000) Fiscal Year 2023: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000) Fiscal Year 2022: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Keywords	speech recognition / Concept Maps / ChatGPT / AI / online learning / concept maps / higher-order skills
Outline of Research at the Start	The research will cover four keys areas: * determine the most effective extended speech maps/Speech Recognition activities and framework, and analyse speech recognition accuracy. * determine whether ECM and Speech Recognition leads to better language outcomes, engagement, self-reflection, and understanding. * (i) refine the teaching framework to be disseminated at teacher workshops, with the aim of allowing teachers to integrate ECMs and Speech Recognition into their classrooms (ii) discover language patterns used by Japanese students. * Attitudes toward ECM creation during Covid-19
Outline of Annual Research Achievements	The purpose of this project is to investigate how the use of Extended Concepts Maps combined with Speech Recognition to improve lower-order and higher-order critical thinking. The first year plan was the technical development of the system and to find student attitudes to the using speech recognition. An initial test into the accuracy of speech recognition was performed and presented at the JALTCALL conference of 2022. It was found that speech recognition was very accurate for native speakers but had varying levels of accuracy with non-native speakers. There was a correlation between speaker level and accuracy but it was not statistically significant due to the small sample and the difficulty ascertaining actual non-native speaker level. A brief pilot study was performed in conjunction with Tokyo Medical and Dental University with Jeanette Dennisson to ascertain attitudes towards using Speech Recognition in a drilled setting and this feedback is still being analyzed but initial analysis shows that the feedback was broadly positive for the use in learning and issues were mainly confined to minor technical issues. Integration of ChatGPT has been performed despite it only appearing in November 2022 with its use being accepted for two future conferences.
Current Status of Research Progress	Current Status of Research Progress 2: Research has progressed on the whole more than it was originally planned. Reason Research and development has progressed as expected. Development of the online system has proved robust, and the accuracy of the system is as expected for online learners. The release of ChatGPT is a game-changer and the researcher has already integrated this technology into his system. Since ChatGPT essentially removes all the weakness of drilled speech practices, this is a major improvement on the core concept of the original proposal.
Strategy for Future Research Activity	The plan for this year is to publish the results of the pilot study as a research paper. Further plans are to ascertain more robustly the use of ChatGPT as a core speech recognition and feedback component within Extended Concept Maps. (This work has already started an a ChatGPT presentation was accepted and completed for a conference in early Fiscal 2023). The preliminary results of these studies and the use of ChatGPT within the system will be presented at EUROCALL 2023 in Iceland, with a paper to be published based on ChatGPT use.