「賴氏人格測驗」為國内廣泛使用與肯定的人格分類工具，然實務上常發現有 些特質難以歸類，再加上其分類標準的統計效果未見討論；因此，本研究欲進一步 運用調整後之標準化殘差（adjusted standardized residual, ASR)法辅助檢視此人格 測驗分類系統於統計上的適切性。本研究以450位大學學生為樣本，在.05與.01統 計顯著水準下，檢視每位受試者在五種人格類型分數得分上的相對差異情形，並交 叉比較分類結果。本研究發現，手冊中被歸為疑問型的人格類型，包含人格特質並 不突出的真正疑問型，以及出現明顯雙重特質的「兼型」人格類型。另外，在各類 型的分類標準中，發現A’、C’、D’等亞型人格分類分數的下限過低，以致有被歸 入F型而彼此區辨不清的現象。文末根據發現並提出上修亞型人格的下限分數、縮 小疑問型的界定範圍、增加「兼型」人格類型的歸類，以及注意内外向人格因素的 解釋等建議。

The Li Personality Test is one of famously applied personality tests in Taiwan. However, no evidences were shown its five-typed criteria supported statistically. Through a nonparametric statistical test (adjusted standardized residual, ASR, approach), the study intends to examine the appropriateness of personality categorized system provided by the Li Personality Test. There were 450 university students participating in the test. Two critical levels of significance, .05 and .01, were used as criteria of classification. Findings showed that "non-identified" personality type can be classified into the "obscured" and the "compatible" ones. Additionally, the lower bound of the classification for subtypes in manual (A’, C’, D’) were too low to distinguished from each other and were dominantly classified as presented with the F type. According to the findings, suggestions of adjusting the lower bounds of personality subtypes, narrowing down the bounds of "obscured" personality types, adding possible "compatible" personality types, and concerning the interpretations of introversive and extroversive personality factors, were provided in the end of text.