Speaker
Description
This JALT Research Grant project examines STEM university students’ emotional responses to English learning using AI‑based text‑mining of reflections collected three times per semester. Joy increased across all classes, while Fear declined mid‑semester and rose slightly at the end. Like steadily decreased, and Sadness briefly disappeared before returning. The first‑year class showed a gradual rise in Sadness, suggesting late‑semester anxiety. These patterns highlight the need for enhanced scaffolding and timely feedback for novice learners.
Abstract section 5: References
Benesse Educational Research & Development Institute, & University of Tokyo. (2021).
高大接続英語教育に関する共同研究報告書（2015–2021）
[Joint research report on English education across secondary and higher education (2015–2021)].
https://benesse.jp/berd/up_images/research/kousaneigo2021.pdf
Brinton, D. M. (2003).
Content‑based instruction. In D. Nunan (Ed.), Practical English language teaching (pp. 199–224). McGraw‑Hill Contemporary.
https://www.academia.edu/38062579/Content_Based_Instruction_A_Relevant_Approach_of_Language_Teaching (academia.edu in Bing)
(academia.edu)
文部科学省. (2021).
全国学力・学習状況調査：英語学習に関する児童の意識調査結果
[National survey of academic ability and learning: Students’ attitudes toward English learning].
https://www.mext.go.jp/
国立教育政策研究所 National Institute for Educational Policy Research. (2023).
全国学力・学習状況調査 報告書（概要版）
[National survey of academic ability and learning: Summary report].
https://www.nier.go.jp/23chousakekkahoukoku/report/data/23summary.pdf
User Local Inc. (2024).
AIテキストマイニングツール ドキュメント
[AI text‑mining tool documentation].
https://textmining.userlocal.jp/
Quirkos. (2024).
Abstract section 2: Contribution/research questions
This study investigates: (1) how student sentiment changes from Week 1 to mid‑semester to Week 16, (2) how content‑based materials influence motivation and emotional responses, and (3) how emotional shifts can be objectively visualized across the semester. The project contributes evidence on evolving affective patterns in English classes and clarifies how CBI tasks may alleviate negative feelings while highlighting late‑semester increases in anxiety that signal the need for additional instructional support.
Abstract section 4: Outcomes/results
Across all four classes, Joy increased steadily, indicating growing enjoyment. Fear peaked in Week 1, declined mid‑semester, and rose slightly in Week 16, reflecting exam‑related anxiety. Like decreased overall, except in one class where mid‑semester motivation briefly improved. Sadness dropped to zero mid‑semester before returning modestly at the end. Anger remained low but increased slightly among students frustrated by stagnant proficiency. English inputs were often misclassified, confirming analyzer limitations. These patterns highlight both the benefits of CBI and the need for targeted late‑semester instructional support.
Abstract section 3: Content/method
Sentiment data were collected at three points: Week 1, mid‑semester, and Week 16 (end‑of‑semester). Responses were written in Japanese, limited to 300 characters, and analyzed using an AI‑based sentiment tool and Quirkos coding. English submissions were planned to be turned into Japanese and then compared using a Japanese‑based User Local app. This approach visualizes emotional trends and identifies how CBI tasks influence learner motivation.
Abstract section 1: Relevance
National surveys show that Japanese students’ positive attitudes toward English decline steadily from elementary through high school, underscoring the growing need for affective support in language learning. This study focuses on STEM students, who often have limited prior English engagement and may experience anxiety when entering university English courses. Content‑Based Instruction can increase relevance and may help reduce motivational decline, yet studies incorporating sentiment analysis remain limited. By examining reflections collected at three points in the semester, this study demonstrates how CBI materials shape emotional responses and motivation in Japanese university contexts.
| Title | Text‑Mining Analysis of Student Sentiment in English Classes |
|---|---|
| Teaching Context | College and university education |