本研究探討熟悉度與可分析性對中文成語理解的影響。實驗一選出不同熟悉度的成語 30個，以 30名大學生為受試者，由受試者寫出每個成語的意義，以測量成語理解的程度。結果顯示：受試者在不同熟悉度的成語上，呈現理解正確度的差異；隨著熟悉度增加，理解正確度隨之提高。實驗二採用 2（熟悉度） ×3（可分析性）實驗設計，選取 36個成語，受試者為 30名大學生，探討成語的可分析性對理解的影響，並檢驗可分析性與熟悉度的關係。結果發現：可分析性具有主效果，並與熟悉度產生交互作用，顯示成語的可分析性可促進成語理解，在高熟悉度條件下，更大幅提升理解正確度。根據研究結果，本文加以討論，並提出成語學習與未來研究的建議。

This study examined the effects of familiarity and analyzability on Chinese idioms comprehension. Experiment 1 used 30 Chinese idioms as materials. Participants were 30 undergraduates and asked to write out the meaning of each idiom. The results showed that familiarity affected participants performances on idiom comprehension, and the comprehension performances were improved incrementally as the level of familiarity increased. With a 2(familiarity) x 3(analyzability) design, Experiment 2 used 36 Chinese idioms as materials, and examined the interaction of familiarity and analyzability. Thirty undergraduates participated in this study. The results showed that the differences in analyzability could influence the comprehension performances in different familiarity levels, and there was a significant interaction of familiarity and analyzability. A small amount of analyzability was enough to trigger more appropriate performance in the low-familiarity condition. With higher level of analyzability, a strong improvement of comprehension performance was found in the high-familiarity condition. The findings suggested that analyzability might play as a role of moderator in the construction of idiom meaning. Some suggestions for future research were provided in the article.