Please use this identifier to cite or link to this item:
Title: 建構英語詞彙知識模型---一個混合模式取向
Constructing an English Vocabulary Knowledge Model---A Mixed Modeling Approach
Authors: 國立臺灣師範大學英語學系
Issue Date: 2008
Abstract: 詞彙知識究竟是建構於單一亦或是多重面向之相關實證結果仍舊處於意見分岐、眾說紛紜的階段。學界目前對於此一議題議論紛紛、莫衷一是。細究原因,此一紛亂不連貫的結果可被歸因於兩個問題。一個是目前學界還未有一個研究能夠準確地校準研究中所用的測驗工具來進而獲得較為穩定的參數估計。另一個問題則是幾乎沒有一個研究是採用驗證性的方式來解決此一紛爭。 因此,為了解決這兩個技術問題,一個結合羅許模式與結構方程模式的混合模組的研究設計將會被提出來嘗試重新檢視詞彙知識的內部建構。羅許模式將被用來校準在研究中所用的測驗工具,而結構方程模式將被用來檢視不同模式的適合度。基於文獻探討的結果,三種不同的模式將分別被提出,相互競逐。總共有四種詞彙知識的測驗工具被使用來建立模式: 1) 詞彙量測驗; 2) 搭配詞測驗; 3) 一字多義測驗及4) 書寫產出誘發性測驗。研究參與者來自於臺灣的大學生。本研究結果同時兼具重要的理論及教學上的意函並且能夠對於詞彙知識內部建構有所更進一步的瞭解。
The empirical results of studies in relation to settling down the issue of unitary/multidimensional vocabulary knowledge are not uniform. The reasons for these mixed and inconsistent findings may be attributed to the fact that none of the existent studies have accurately calibrated the test instruments used in breadth/depth vocabulary studies and that few studies have taken a confirmatory approach to solving the debate. To address the lacuna in the literature, a mixed modelling approach – a mixture of Rasch modelling and structural equation modelling – will therefore be used in sequence in the study. Rasch modelling will be used for calibrating test instruments of the study, and structural equation modelling will be used for testing the fit of proposed models. Three kinds of rival models, based on the review of literature, are respectively proposed for empirical testing. There are in total four types of vocabulary knowledge measures to be used for modelling: 1) vocabulary size test, 2) collocation test, 3) polysemy test, and 4) prompted productive written form test. Participants of the study will be Taiwanese college students. The current study has both important theoretical and pedagogical implication. The results of the study will advance and deepen our understanding toward the construct of vocabulary knowledge.
Other Identifiers: ntnulib_tp_B0218_04_002
Appears in Collections:教師著作

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.