以Kano二維模式及資料探勘分析大學土木系審查重點及學生入學前後表現

Abstract

大學多元入學實施以來，發展符合校系特色之招生策略與入學審查機制為各校系首要課題。故本研究旨在探討各大學土木系招生入學審查之需求、學生特質與後續表現之相關性。提出Kano二維品質模式於大學土木系入學審查之可行性，並以交大土木系為例，運用資料探勘技術之關聯規則與群集分析探討(1)學生背景與大學成績潛在關係；(2)土木系學生轉系影響因子；(3)名校效益與名次效益；(4)大學入學前後成績關聯；(5)土木系各必修科目間成績關聯。提供將來科系擬定招生策略之參考依據。 本研究結果顯示交大土木系學生之性別與高中類別的不同確實會影響其大學成績表現；近親從事土木業之學生較不會轉系，但並不表示未轉系之學生其近親皆從事土木業；至於名校效益與名次效益本研究發現此現象是存在的；且發現各必修科目成績間之關聯規則。此外，該系之書面審查分數較學測採計分數更能預測學生大一成績之表現。

The aim of this research is to understand relationship among pre-and post- entrance performance of department of civil engineering students, and entrance evaluation criteria. Put forward the feasibility of kano model in entrance evaluation criteria. Apply association rule and cluster analysis to explore: (1) the relationship between student background and college grades, (2) the influencing factors of changing major, (3) prestigious school and place benefits; (4) the relationship among pre-and post- college entrance scores, (5) the relationship among the compulsory subject score of civil engineering. We found the difference of gender and high school categories would affect students’ college grades. Students were less likely to change their major, if their next of kin engaged in civil engineering industry. Then, different academic achievement or different graduated schools would affect college students’ academic performance. We also found association rule among the compulsory subjects. Moreover, Document review is more correct than SAT score, when predicting first-year college grades.