本研究目的有三：一、探討雙語使用者的詞彙與概念的聯結方式，即考驗「概念媒介論」及「文字連結論」兩種語言詞彙表徵模型何者可成立。二、比較不同外語能力受試者的詞彙與概念之連結方式的差異，亦即考驗「中間論」的預測是否成立。三、檢驗「相減法」及「促發法」兩派典對語言詞彙表徵理論之設考驗結果的一致性。 實驗－以20名「英語能力非熟練組」大學生為受試，利用反應時間的「相減法」配合中文詞及圖形的命名(翻譯)或分類作業進行實驗。結果發現：中文詞與圖形的英文命名反應時間之差小於中文詞與圓形的中文分類反應時間之差。「文字連結論」的預測獲的得支持。 實驗二以20名「英語能力熟練組」大學生為受試，實驗方法與實驗－相同。其實驗結果發現：中文詞與圖形的英文命名反應時間之差小於中文詞與圖形的中文分類反應時間之差。此結果與實驗－發現相同，亦支持「文字連結論」的預測。 實驗三以24名大學生為受試者，利用「促發法」配合字(詞)彙判斷作業，比較語言內相關字、語言間翻譯字、及語言間相關字等三種促發項對目標項的促發效果。實驗結果發現：中文目標項並未具有明顯的促發效果，因此未能支持本研究中有關促發效果關係的假設。在英文目標項部份，則發現具有明顯不同的促發效果型態，且語言間重複促發效果相近於語言內語意/連結促發效果，而此兩種促發效果又均大於語言間語意/連結促發效果。此種促發效果型態支持「文字連結論」的預測。

　　　　　The purposes of this thesis were: (1) to explore the lexical and conceptual representations of Chinese-Englis bilingual university students, namely, to test the 'concept-mediation hypothesis' and the 'word-association hypothesis'; (2) to compare the patterns of lexical and conceptual representations of bilingual university students with different levels of English proficiency, namely, to test the 'intermediate hypothesis'; (3) to compare the 'subtraction method' paradigm and the 'semantic/associative priming' paradigm in testing the above hypotheses. Twenty English non-proficient subjects participated in experiment 1. They were required to perform in two kinds of tasks: Chinese words translating, pictures naming (in English), and words and pictures categorizing (in Chinese). The subtraction method paradigm was used. The results showed that the reaction time difference between words categorizing and pictures naming is significantly smaller than that of words categorizing and pictures categorizing. These results supported the prediction of 'word-association hypothesis'. In experiment 2, twenty English proficient subjects were asked to perform in the same tasks as in experiment 1. The results were in line with those of experiment 1, also supporting the prediction of 'word- association hypothesis'. Twenty-four subjects participated in experiment 3, adopting the 'semantic/associative priming' paradigm along with lexical decision task. The results showed that the priming effects were not significant when Chinese target words were used. However, the priming effects were significant when English words were used as target words. Since 'between-language repetition effect' was equal to 'within-language semantic/associative priming effect', and both were greater than 'between-language semantic/associative priming effect', the prediction of 'word-association hypothesis' was again supported.