本文論的特色有三；在研究過程中：(1)提出以較普遍、簡易且使受訪者回答意願較強的複選式類目資料勾選的調查方式蒐集到不失詳細且較真確的資料；(2)在資料分析方面，藉著交互應用Correspondence Analysis (CA) 與Cluser Analysis 的優點及特色，克服了以往只能簡單分析多變量類目資料等的限制，而充分、完整地量化類目資料的特質及其關係結構；(3)並且以簡單、清晰、最為大眾接受的適圖形解說結果。實例中搭配應用CA與K-means Cluster Analysis的特色，提出一種與以往不同的學生人格特質區隔分群過程不但能將受訪學生依據其所選擇人格類別項目上的異同在不漏失資料的情況下區隔分群，並將使用階層集群技巧（Hierarchial grouping techniques）的連結關係增加了CA二維平面圖的解析力。經由實証分析結果發現本校不同科系學生的人格特質各具特色，經由過本研究實証的新經驗，除了能與校內學生輔導中心的學生個人之人格特質資料進行有效整合及比對之外，並且能提供予校內訓輔、職輔等單位以及各科系決策者在考量配合學生的人格特質傾向進行輔導性課程調或專題活動安排時之有效參考。

There are three features in this paper. The proposes (1) provide the way of using multiple-responses-categorical-data survey instrument, which is more popular, simpler, and to which the interviewees are more willing to answer, to obtain competitively detailed and more reliable data; (2) in respect of data analysis, by alternately applying the advantages and features of correspondence analysis (CA) and custer analysis, sufficiently and completely quantify the characters and relation of categorical data without the former restriction of only simply analyzing multistage categorical data; and (3) explain the results using optimal graphical displays, which are simple, clear and most acceptable to the public. Applying complementary use of CA and K-means cluster analysis, this paper features a student-character-segmentation process, which is different from the former ways, and which is not only able to segment interviewed students, without losing data, according to the difference and sameness between the character categories they chose, but also uses the CA map to describe the relation between and within obtained groups and chosen categories, etc., and meanwhile which uses hierarchical grouping techniques to enhance the joint space map obtained from 2-way correspondence analysis. Further, the examples provide a clear and more reliable reference for education improvement.