When one is Not Enough: Translation Rating and the Assessment of Partial Word Knowledge

當一種方法不足時：字彙譯解及部分知識之評量

杜凱蕾(Carol Troy) ； 曹秀蓉(Syou-Rung Tsau)

《人文暨社會科學期刊》 6卷1期 (2010/06) Pp. 87-108

https://doi.org/10.7118/JHSS.201006.0087

摘要

在非刻意字彙習得研究裡，對於部分字彙知識作正確的評量是重要且必需的，開放式的母語譯解愈來愈常被用來作此類的評量，何種精確度適用於字彙譯解評量？為了回答這個問題，我們分析一份「字彙譯解前測→閱讀→字彙譯解後測」的實驗數據。我們使用11點量表評量字彙譯解前測與後測。接著經由一個概算程序，我們得到對應的2點、3點及6點量表資料。我們採用曼惠特寧U法檢定這四組資料（即11點、2點、3點及6點）以便找出受測者在前後測的字彙知識進步達到顯著水準的那些單字。以最原始的11點量表為標準，我們發現2點及3點量表造成錯誤的正數及負數。依此我們得到兩個結論：1.並非所有部分正確的字彙譯解可以得到相同的分數；2.多點量表比2點及3點量表能更準確評出字彙譯解的成績。我們進而探討實用的評量方法及使用前後測比只使用後測有更多的優點。

關鍵字

非刻意字彙習得 ； 專業英語 ； 部分字彙知識 ； 譯解評量 ； 字彙評量

並列摘要

An important requirement in foreign language incidental vocabulary acquisition research is accurate vocabulary assessment of partial word knowledge. Open-format L1 translation tests are increasingly used for this purpose. What level of precision is appropriate in the translation rating procedure? To answer this question, experimental data from a read-and-test study are analyzed. The pretest and posttest translations are rated on an eleven-level scale. Through an approximation process, equivalent binary, three-level, and six-level data are derived. The Mann-Whitney U Test is applied to each of the four data sets (eleven-level, binary, three-level, and six-level) to identify the words for which subject knowledge improvement reached significance. By using the original, eleven-level data as a standard, it is shown that binary and three-level rating lead to false positives and false negatives. Two conclusions are drawn: 1. Not all partially correct translations deserve equal credit; and 2. Multi-level rating is a more precise measure of translation accuracy than binary and three-level rating. Practical rating issues and the advantages of using a pretest and posttest as opposed to a posttest only are also discussed.

並列關鍵字

incidental vocabulary acquisition ； partial word knowledge ； translation rating ； vocabulary assessment

參考文獻

(Ferris, D. (1988). Reading and second language vocabulary acquisition. Unpublished manuscript, Department of Linguistics, University of Southern California.).

(Meara, P., & Jones, G. (1990). Eurocentres vocabulary size test (version E1.1/K 10, MSDOS), Zurich: Eurocentres Learning Service.).

Watts, J. (2002). The effect of multiple texts on vocabulary acquisition. Retrieved September 15, 2009, from http://www.tesolanz.org.nz./

Anderson, R.,Freebody, P.(1983).Effects on text comprehension of differing proportions and locations of difficult vocabulary.Journal of Reading Behavior.15(3),19-39.

Beglar, D.(2000).Estimating vocabulary size.JALT Testing & Evaluation SIG Newletter.4(1),2-3.