歡迎來到文藻W-Portfolio

論文名稱	Exploring College Student English Proficiency Test (CSEPT) Washback to Taiwanese EFL College Students in General English Classes.
研討會開始日期	2020-05-29
研討會結束日期	2020-05-30
所有作者	Chiu, Yi-Min, Cheng, Richard C. H., Lo, Yi-Jou, Wu, Rachel Yi-fen, & Lee, Lydia Hsin-Yun.
作者順序	第二作者
通訊作者	否
研討會名稱	the 37th International Conference on Language Teaching and Learning & the 22nd International Conference and Workshop on TESL and Applied Linguistics
是否具有對外公開徵稿及審稿制度	是
研討會舉行之國家	NATTWN-中華民國
研討會舉行之城市	桃園
發表年份	2020
所屬計劃案	無
可公開文檔
可公開文檔
可公開文檔

[摘要] :
This paper is part of a one-year-long joint research project between an English language
teaching center at a four-year technical university and the Language Training and Testing
Center (LTTC) in Taiwan. The former has been using College Student English Proficiency
Test (CSEPT, level 2) designed by the latter as entry placement test, annual progress
assessment, and graduate exit exam for students enrolled in the General English (GE)
program. The project has multiple purposes. Yet, this paper solely explores the relationship
between internal GE exam and external CSEPT test. The GE program indeed does not narrow
its focus to test preparation and thereby not aligned to the high-stakes CSEPT test. It remains
unclear, however, if the two tests under comparison are similar in their content and how this
finding may explain students’ CSEPT score gains.
Participants include 142 Taiwanese EFL college freshmen, placed into seven GE classes
at two different levels based on their CSEPT scores upon entry in fall 2019. Collection of
qualitative and quantitative data lasts for two semesters between fall 2019 and spring 2020.
Qualitative data comprise reading texts and comprehension questions from two levels/sets of
four internal GE (two midterm and two final) exams and three external CSEPT tests (the
entry placement test as pre-test, annual assessment test as post-test, plus an additional test).
Adopting content analysis or one 15-item/feature and one 2-item/feature proformas, invited
experts analyze reading texts and questions respectively in each GE exam and those in each
CSEPT test. An additional CSEPT test is included in the content analysis in order to
investigate whether contents across CSEPT test forms are parallel in terms of criterial
features. Quantitative data consist of students’scores on the CSEPT pre- versus post-tests and
undergo statistical analysis using paired t-tests. Also presented are preliminary findings and
pedagogical suggestions.

[英文摘要] :
This paper is part of a one-year-long joint research project between an English language
teaching center at a four-year technical university and the Language Training and Testing
Center (LTTC) in Taiwan. The former has been using College Student English Proficiency
Test (CSEPT, level 2) designed by the latter as entry placement test, annual progress
assessment, and graduate exit exam for students enrolled in the General English (GE)
program. The project has multiple purposes. Yet, this paper solely explores the relationship
between internal GE exam and external CSEPT test. The GE program indeed does not narrow
its focus to test preparation and thereby not aligned to the high-stakes CSEPT test. It remains
unclear, however, if the two tests under comparison are similar in their content and how this
finding may explain students’ CSEPT score gains.
Participants include 142 Taiwanese EFL college freshmen, placed into seven GE classes
at two different levels based on their CSEPT scores upon entry in fall 2019. Collection of
qualitative and quantitative data lasts for two semesters between fall 2019 and spring 2020.
Qualitative data comprise reading texts and comprehension questions from two levels/sets of
four internal GE (two midterm and two final) exams and three external CSEPT tests (the
entry placement test as pre-test, annual assessment test as post-test, plus an additional test).
Adopting content analysis or one 15-item/feature and one 2-item/feature proformas, invited
experts analyze reading texts and questions respectively in each GE exam and those in each
CSEPT test. An additional CSEPT test is included in the content analysis in order to
investigate whether contents across CSEPT test forms are parallel in terms of criterial
features. Quantitative data consist of students’scores on the CSEPT pre- versus post-tests and
undergo statistical analysis using paired t-tests. Also presented are preliminary findings and
pedagogical suggestions.