臺灣大學教學成果與授課型態關係之探討

截至2008年為止，台灣的學院與大學的數量已經增加到了162所，在這樣快速擴張之下，可能會降低台灣高等教育的品質。然而，教學與研究是大學的兩個主要功能，為了鼓勵教職員提高其教學品質，國立台灣大學每年都頒發教學傑出獎以及教學優良獎給予這些優秀的教師。而本研究收集了包括得獎紀錄、學院別、性別、職別、授課平均人數、有無教授大班課程(>100人)、教授大班課程(>100人)的學期數、有無教授研究所考試所需的課程、教授研究所考試所需課程的門數，以及是否只教授研究所以上的課程等資料，來探討其授課型態與教學成果的關係。本研究的研究對象為國立台灣大學在2003到2007年間皆為專任之教師。我們利用卡方檢定來比較兩個比例是否有差異，再利用Cochran-Armitage檢定來看比例之間的趨勢，而Mann-Whitney U統計量是用來檢定非類別的連續型變數，最後再用邏輯斯迴歸來看，所有授課型態的自變數與得獎機率之間的關係。在單變量分析中顯示，得獎機率與學院別、性別、職別、授課平均人數、有無教授大班課程(>100人)、教授大班課程(>100人)的學期數、有無教授研究所考試所需的課程、教授研究所考試所需課程的門數，以及是否只教授研究所以上的課程等自變數間，皆有統計顯著相關，然而，利用邏輯斯迴歸進行整體分析時發現，只有性別、授課平均人數與教授研究所考試所需課程的門數，三個自變數與得獎機率有關。總而言之，在國立台灣大學專任教師中，若為女性之專任教師且授課平均人數較高並敎授越多門研究所考試所需之課程者，其得教學傑出獎或教學優良獎的機率會相對的提高。

關鍵字

教學成果；授課型態；教學品質；卡方檢定；邏輯斯迴歸

並列摘要

The number of colleges and universities in Taiwan has increased to 162 in 2008. Such a rapid expansion in quantity may results in a low quality for Taiwan’s higher education. However teaching and research are two primary functions of universities. To encourage the faculty to improve the quality of teaching, the National Taiwan University has given out the Excellent Teaching award and Outstanding Teaching award each year. To investigate the association of teaching patterns and teaching performance, we collected the data of receiving awards, institutes, gender, ranks, average class size, class size greater than 100 students, the semester number of class size greater than 100 students, the course taught required by entrance examinations of the graduate school, the number of the courses taught required by entrance examinations of the graduate school, and whether a faculty only teaches graduate-level courses. Our targeted population is the full-time faculty employed by the National Taiwan University during the entire period between 2003 and 2007. We used chi-square test for comparing two proportions, Cochran-Armitage test for trend in proportions, Mann-Whitney U statistic for non-normal continuous data, and logistic regression for the relationship between probability of receiving the award and independent variables of teaching patterns. Univariate analyses revealed that the probability of receiving the award is statistically significantly related with institutes, gender, ranks, average class size, class size greater than 100 students, the semester number of class size greater than 100 students, the course taught required by entrance examinations of the graduate school, the number of the courses taught required by entrance examinations of the graduate school, and whether a faculty only teaches graduate-level courses. However logistic regression found that the probability of receiving the award is only related to gender, average class size, and the number of the courses taught required by entrance examinations of the graduate school. In summary, a NTU faculty who is female and teaches bigger classes and large number of courses required by entrance examinations of the graduate school will have a high probability receiving an Excellent Teaching award or Outstanding Teaching award.