過去的文獻中指出一些與教學品質無關之因素,卻對學生評鑑教師之教學評鑑分數有顯著的影響,導致教學評鑑分數之不公。朱怡潔(2007)曾經探討過,如何利用迴歸模型,消除「學制別」、「必選修別」、「學院別」、「班級人數」、「學生預期成績」等五個外在干擾因素造成教師教學評鑑分數不公的現象,惟班級人數變數未作全面之考量。本研究主要依循朱怡潔(2007)文中教學評鑑模型,對班級人數變數設定方式作整體考量及調整,重新修正教師教學評鑑模型。而模型的殘差部分為過濾了與教學品質無關的干擾因素之數值,故以殘差分數發展為評鑑教師教學表現的新指標。 以修正後教師評鑑分數為基礎,進行各項多變量分析,由不同角度探討教師教學表現之差異。利用因素分析、集群分析、多元尺度分析等多變量分析方法,區分教師教學特色、深入探討教師教學表現之差異及教師教學表現之定位。又以修正後教師評鑑分數,觀察教師教學評鑑分數分布與變異情形,可提供教師檢視自我教學品質管理的好壞,亦能作為系所排課參考用途。
Many studies have pointed out that teaching evaluation by students is affected by some factors which are unconcerned with teaching quality and make the evaluation unfair and not applicable. The purpose of this study is to construct a regression model that can provide a fair score on teaching evaluation. Regression models of raw teaching evaluation score on five interfering variables, “college”, “graduate/undergraduate program”, “required/optional course”, “class size” and “expected grade” are built up. Excluding the effects of the interfering variables, the residuals of the models are used as the new measures for adjusting the score of teaching evaluation. The adjusted score of teaching evaluation is further used to describe and compare the teaching performance among teachers. Factor analysis, cluster analysis and MultiDimentional Scaling are applied to describe and distinguish instructors teaching characteristics. A teaching effectiveness and variation analysis for the instructors is also proposed. The results of the analysis will help instructors and school in course arrangement and further consulting programs in teaching.