2023 Fiscal Year Research-status Report
Investigating Multiword Sequence and Speaking Fluency Influences in High-stakes Assessments
| Project/Area Number |
22K00700
|
|---|
| Research Institution | Hiroshima University |
|---|
Principal Investigator |
Hougham Daniel 広島大学, 外国語教育研究センター, 准教授 (10829352)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
CLENTON JONATHAN 広島大学, 人間社会科学研究科(総), 准教授 (80762434)
内原 卓海 東北大学, 国際文化研究科, 講師 (10905847)
Higginbotham George 叡啓大学, ソーシャルシステムデザイン学部, 准教授 (20885090)
|
|---|
| Project Period (FY) |
2022-04-01 – 2025-03-31
|
| Keywords | multi-word sequences / lexical bundles / learner-corpus research / oral fluency / oral proficiency / multiple regression / random forests analysis / dominance analysis |
|---|
| Outline of Annual Research Achievements |
We have completed the second phase of analyzing the relationship between lexical bundle (LB) use and oral fluency. We identified 119 frequent LBs, mostly three- and four-word units. One-way ANOVAs indicated significant differences in four LB measures across proficiency levels (IELTS 6.5-7.5). Employing robust regression, dominance analysis, and random forests, we found that bigram mutual information (MI) strongly predicts higher proficiency levels. We also discovered that longer LBs enhance speed fluency and reduce mid-clause pauses, highlighting their unique processing efficiencies. Conversely, shorter LBs, particularly those with high mutual information scores, subtly yet positively influence articulation rates and reduce speech repairs, highlighting their key role in fluency dynamics.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
To date, we have presented our results at seven international conferences, including the recent Vocab@Vic. We published a pilot study in a top-tier journal and have two manuscripts currently under review. We have also collected additional data which will allow us to conduct new analyses across five proficiency levels, bringing our total sample size to upwards of 180 participants. However, the increase in participants was less than expected, with fewer at the lowest and highest levels compared to the middle levels. Having been busy with manuscript writing, we have not yet processed our additional data for fluency measurements.
|
| Strategy for Future Research Activity |
In our future research, we plan to: (a) Collect additional data at the lowest and highest levels (IELTS 6 and 8), in order to balance our groups for more robust analyses across five proficiency levels. Our next data collection is tentatively scheduled for September 2024; (b) Investigate the LB-fluency relationship with a larger sample size across five proficiency levels (IELTS 6–8); (c) Explore additional quantitative analyses (e.g., potential influence of L1 background on LB usage and proficiency scores); (d) Explore qualitative analyses (e.g., presentation topics; LB usage variation based on functional classification); (e) Present at Vocab@Vic 2025 (in Maryland, USA); (f) Prepare an additional manuscript for submission to another top-tier journal.
|
| Causes of Carryover |
I plan to travel abroad to the location (Queen Mary University of London) where our continued data collection will take place. We plan to continue processing our data set for fluency measures with the help of research assistants. We also plan to present our results at an international conference in the near future. The reason why some funds remained in FY2023 is that we did not have enough time to hire and train an additional research assistant to help with processing our dataset for fluency measures.
|
