採自陳式量表之實證結果再檢視–考量共同方法變異

Abstract

在行為科學的研究之中，研究者為了瞭解潛在構念間相關之現象，往往透過「自陳式（self-report）」量表衡量構面並進行分析研究。但若於研究中資料收集過程使用單一來源之方法，會使得量表分數受到測量方法或受測者本身之因素受到影響，那麼這個研究即會出現研究方法中所謂的「共同方法變異（common method variance；CMV）」問題。 共同方法變異（common method variance；CMV）問題將會嚴重導致實證研究之結果有所干擾，且將產生潛在性且具有誤導性之結論(Campbell & Fiske, 1959)。共同方法偏誤問題將會不當地導致構念間的相關係數膨脹，產生型一誤差（type I error）以致於相關不當累積，出現不當的結論；卻也有可能減弱了變項間的關係，產生型二誤差（type II error）以致於錯失了原本該顯著的構念相關。 於此，本研究以中央研究院調查研究中心所提供之研究計畫進行實證分析，收集了於管理類別中之未考慮共同方法偏誤問題之計畫進行分析。而本研究針對所收集之計畫進行共同方法偏誤問題之矯正後，發現考量之前後於假說之驗證尚有重大之改變。

In the research of behavioral sciences that aimed to find the relationship between latent constructs, most researchers used the self-report scale to measure and analysis. However, the data that collected from single source may cause the score of the scale will be influenced by the respondent himself, which will cause common method variance. Common method variance will cause a serious problem of the research, and it will generate a misleading conclusion (Campbell & Fiske, 1959). The Common method variance can either inflate or deflate observed relationships between constructs, thus leading to both Type I and Type II errors. For this purpose, the study uses the projects from Survey Research Data Archive in Academia Sinica and to re-analyze the effect by adding common method variance in project analysis. By re-analyze the projects in Academia Sinica, these findings challenge the conclusion of the original project which did not consider common method variance problem.