2021 Fiscal Year Research-status Report
Measuring the extent speech-to-text functionality aids speaking fluency (breakdown & speed-fluency) in EFL learners using a mobile assisted language learning app
| Project/Area Number |
19K00803
|
|---|
| Research Institution | Teikyo University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
ボビー ヒロユキ 九州産業大学, 語学教育研究センター, 准教授 (20536247)
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | speed fluency / breakdown fluency / speaking fluency |
|---|
| Outline of Annual Research Achievements |
The pilot data collection phase was completed and analysis showed that there was a significant improvement in breakdown fluency (reduction in the number of pauses per unit of speaking time). This translated into smoother speech patterns and improved perceived fluency by native raters. The 5 hour treatment group was inconclusive, but the 10 hour treatment group showed a statistically significant improvement with a moderate effect size. The speaking speed fluency showed similar trends in both the 5 hour and 10 hour treatment groups over the control group. The 20 hour treatment group data had an unusually large number of outlier data points that were difficult to explain and might have been due to data corruption or a failure to follow protocols that were given. We are planning on recollecting the 20 hour treatment data along with some additional control data for a more robust analysis.
Pushed output by the speaking/shadowing treatment is hypothesized to be aided by several other features in the data acquisition system. These included the comprehension feature, the vocabulary activity, and the grammar building activity. A more rigorous experimental design might be needed to accurately determine the effects of each component activity on breakdown fluency and speed fluency of the main treatment. In order to compare this data with the control data, we have included similar protocols for all groups including the control participants. We will continue to analyse this aspect of the experiment this year.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
The pilot data collection phase has been completed and we have moved on to the main data collection phase and have started preliminary analysis of the data and related metrics. We have collected data on the 5 hour and 10 hour treatment group but have not completed the 20 hour treatment group due to data corruption and an unusually large number of unaccounted for data outliers. We will restart that element of the research in the first half of this last year of the research project. We may attempt to collect additional 5 hour and 10 hour treatment data on participants at varying levels of English ability in addition to the 20 hour treatment group and control group. This will give a more robust final analysis. We will continue along these lines. The effects of other metrics such as session duration, score and completion rates are being looked at as to whether they influence any improvement or reduction in speaking breakdown fluency and improvement in speaking speed fluency. These are being included in the statistical analysis as covariants and should make the investigation more robust. Furthermore, we are looking into acquiring more reliable control data by adjusting the control protocols within those participant groups. If this has a positive outcome, we will potentially have a statistically larger effect size by eliminating several covariates.
|
| Strategy for Future Research Activity |
We are going to continue to collect data in the first semester of the last year and may collect some supplemental data in the fall semester depending on how data acquisition proceeds. Disruptions due to Covid have been reduced and we plan on finishing collecting the main phase of the research data. Unfortunately, due other institutional changes of the primary researchers there have been some delays in the progress of the overall plan. However, the intent is to complete the main data collection this year. The 20 hour treatment group is planned this year with additional control groups with autonomous learning and difficulty adjustment integrated into the data collection.
We have started to disseminate our research findings by presenting at several international and domestic conferences and will continue to do so this year and into the next. Several papers are planned for this year and we hope to continue with the analysis of the data into the beginning of 2023.
|
| Causes of Carryover |
Covid prevented planned international conference attendance. We have started to disseminate our results at several online and domestic conferences and will start traveling internationally to conferences outside of Japan this year.
|
