近義詞教與學是近來華語教學研究中一項重要的議題，對華語學習者而言，近義詞辨析一直是詞彙學習難點之一。而收集大量學習者語料的中介語語料庫有助於探究華語學習者的詞彙使用偏誤情形，可作為近義詞教學時的重要參考資料。本文以國立臺灣師範大學建置之華語學習者語料庫為語料來源，調查兩組常用動詞的混用現象：「幫」、「幫忙」、「幫助」及「變」、「變得」、「變成」。研究結果發現「幫」取代「幫忙」、「幫助」及「變」取代「變得」、「變成」的偏誤情形在語料中比例較高，而其偏誤皆主要表現在句法結構的混淆。此外，本文亦進一步檢視近義詞組於華語教材中的呈現情形，嘗試釐清教材與學習者偏誤的關聯。期能藉由分析結果提供華語近義詞教學更多的參考。

Synonym teaching has been one of the important topics in teaching Chinese as a second language in recent years. A Chinese learner corpus can help researchers better identify common synonym errors, and the results can further facilitate the teaching and learning of synonyms. In this study, two sets of synonymous verbs: “bang”, “bang-man”, “bang-zhu”, and “bian”, “bian-de”, “bian- cheng” were investigated based on the Chinese learner corpus developed by National Taiwan Normal University. The results showed that the replacements of “bang-man”, “bang-zhu” by “bang” and that of “bian-de”, “bian-cheng” by “bian” were the most common learner errors, which were mainly attributed to learners’ confusion of the sentence structures. In addition, the possible connections between the presentation of these verbs in the teaching materials and the learners’ misuses of these verbs were also discussed in this study. We expect that the findings of this study will be useful for teaching Chinese synonymous verbs.