PBT-TOEFL與CET-6考試閱讀測驗之比較性研究

Investigating the Comparability of the Reading Sections of Two English Proficiency Tests: PBT-TOEFL and CET-6

董炤靈(Chao-Ling Tung)

指導教授 ： 劉蕙君

中原大學/人文與教育學院/應用外語研究所/碩士(2010年)

https://doi.org/10.6840/cycu201000012

摘要

摘要

本論文在探討PBT-TOEFL和CET-6考試中閱讀測驗的相關性。在本研究中，同期效度，建構效度，信度，試題難易度及鑑別度研究用來提供更全面性的相關性評估。此外，藉由收集受試者背景資料的簡易問卷也能當作之後分析與討論的參考。在本研究中，107位受試者是來自於台灣北部某一所大學英文系的學生。當中，根據他們在這兩項考試的表現，被分為前27％的高分組和後27％的低分組，以當作分析試題鑑別度的分組標準。此外，為了進一步分析試題，3位曾經受過語言學相關課程訓練的專家參與試題類型的分類。本研究使用的兩份考題 (2005年1月份的試題)在一天當中的90分鐘內完成。本研究之主要發現如下: 首先，研究發現考生在PBT-TOEFL和CET-6閱讀測驗之表現呈現正相關；對這107位受試者而言，PBT-TOEFL和CET-6閱讀測驗分數之間有相當密切的相關性。另外，因素分析結果顯示，一般性的閱讀能力被歸納為PBT-TOEFL和CET-6閱讀測驗主要的共同測試能力。再者，研究發現PBT-TOEFL和CET-6閱讀測驗之間有相似的信度係數。最後，試題難易度和鑑別度的相近數值結果更進一步為這兩項測驗相關性研究提供進一步的證據。

關鍵字

試題相關性 ； 效度 ； 信度 ； 英語能力測驗 ； 閱讀測驗

並列摘要

Abstract

This present study aims to investigate the association between the reading sections of Paper-Based Test of English as a Foreign Language (PBT-TOEFL) and College English Test Band-6 (CET-6). In this study, research issues of concurrent validity, construct validity, reliability, item difficulty and item discrimination between these two tests are investigated in order for a more comprehensive assessment on the level of association. Subjects participating in this study are 107 English majors from a college in the northern Taiwan. Based on their performance scores on both tests, the test takers were classified into high and low proficiency groups---the high proficiency group consisting of those whose scores at the top 27％ and the low proficiency group consisting of those whose scores at the bottom 27％. Moreover, three language experts were asked to classify item types, which were used for item analysis. Two tests in the form of selected past papers, 2005, 1, were distributed to study participants within 90 minutes in one testing date. The main findings of this study are listed as follows: Firstly, there is positive correlation between the reading sections of PBT-TOEFL and CET-6. It is concluded that scores of PBT-TOEFL and CET-6 display close association for these 107 participating subjects. Secondly, through the factor analysis, the general English reading proficiency is identified to be the common construct underlying these two tests. Thirdly, results of the statistical analyses show that the reliability levels of these two tests are similar. Finally, comparable item difficulty and item discrimination of the two tests provide further evidence for the association between these two tests.

並列關鍵字

test comparability ； reliability ； validity ； English proficiency tests ； reading assessment

參考文獻

Brown, J. D. (1999). Relative importance of persons, items, subtests and languages to TOEFL test variance. Journal of TLanguage TestingT. 16(2). 216-237.

Cho, Y. (2002). Advanced Placement International English Language (APIEL). Journal of Language Testing. vol. 19: 323-331.

Choi, I. C. (2003). Comparability of a paper-based language test and a computer-based language test. Journal of Language Testing. 20(3). 295-320.

Davies, A. (1990). Principles of Language Testing. Oxford: Basil Blackwell.

Ebel, R. L.,&Frisbie, D. A., (1991). Essentials of Educational Measurement.