2014 Fiscal Year Research-status Report
Mobile system to monitor extensive reading
| Project/Area Number |
25370643
|
|---|
| Research Institution | Fukuoka Women's University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
W・R PELLOWE 近畿大学, 工学部, 講師 (30330327)
LAKE J 福岡女学院大学, その他部局等, 講師 (60610235)
|
|---|
| Project Period (FY) |
2013-04-01 – 2016-03-31
|
| Keywords | Extensive reading / Reading difficulty / Vocabulary level / Text analysis / Rasch analysis / 読書の利益 / Reading gains |
|---|
| Outline of Annual Research Achievements |
Data from the MOARS Extensive Reading module was gathered during 2014 and analyzed using many-faceted Rasch measurement (MFRM) to find the difficulty of popular books. Students were given diagnostic tests of vocabulary size, morphology, and reading speed.
Samples of text from 380 popular graded readers were analyzed to give Lexile measures of difficulty (Stenner, Burdick, Sanford, & Burdick, 2007), and also readability measures using the Flesch-Kinkaid grade level and Flesch Reading Ease scales. This took longer than expected.
Preliminary analysis focused on validating the diagnostic tests. One research paper has been published (Holster & Lake, 2015) and a second paper is under final review (Holster & Lake, forthcoming). Next, we investigated the relationship between text features and text difficulty. The Lexile Framework provided better predictions of text difficulty than the Flesh-Kinkaid grade level and Flesch Reading Ease scale. However, the Lexile framework, which was developed for native speaker readers, uses sentence length and vocabulary frequency to estimate text difficulty. Our results showed that reading difficulty related to sentence length, but not to vocabulary frequency. Also, the length of the book was a very strong predictor of difficulty. These findings suggest that the frequency of vocabulary in native-speaker corpuses is not useful for predicting the difficulty of texts for second-language learners and the extensive reading programs should focus on reading many short books instead of setting word-count targets.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
Data collection from the MOARS system was trouble-free and provided a very large quantity of data about student reading ability and book difficulty. The analysis of the scanned text samples took longer than expected and was not finished until March 2015. One co-researcher (Lake) moved to Fukuoka Jogakuin University in April 2014 and the principal researcher (Holster) moved to Fukuoka University in April 2015. The collection and analysis of data is now on schedule again, but the reporting of results has been delayed by about six months because of the delays.
|
| Strategy for Future Research Activity |
The main focus in 2015 will be writing research reports on graded reader difficulty. Data collection is complete, and three research papers are planned for completion by March 2016. One paper will report on the relationship between reading difficulty and text features such as vocabulary frequency and sentence length, a second paper will report on the vocabulary profiles of graded readers, and the final paper will report on the measurement of reading speed.
Preliminary results will be reported at the Foreign Language Education and Technology conference in Boston, U.S.A. in August 2015, the Pacific Rim Objective Measurement Symposium in Fukuoka, Japan in August 2014, the日本言語テスト学会 in September 2015, and the JALT International Conference in Japan in November 2015.
|
