查詢結果分析
來源資料
頁籤選單縮合
題名 | 多變量變異數分析的顯著性考驗=Multivariate Test Statistics in Multivariate Analysis of Variance |
---|---|
作者 | 傅粹馨; Fu, Tsuey-shing; |
期刊 | 教育研究 |
出版日期 | 19970800 |
卷期 | 5 1997.08[民86.08] |
頁次 | 頁1-14 |
分類號 | 319.5 |
語文 | chi |
關鍵詞 | 多變量變異數分析; Multivariate Analysis of Variance; |
中文摘要 | 多變量統計方法於研究中之重要性有三:一、控制實驗錯誤率之膨大;二、反映 社會的真實面; 三、 同時考量依變項間之關係。 多變量變異數分析 (multivariate analysis of variance; MANOVA) 是多變量統計方法之一, 其顯著性之檢定有四: Wilks' lambda (A), Pillai's trace (V), Hotelling trace (T) 和 Roy's greatest root (θ ) 。 用 SAS 與 SPSS 執行 MANOVA 時Λ,V,T, θ均列於報表中。本文內容包括:一、特徵值 在 MANOVA 中所扮演的角色;二、MANOVA 之顯著性考驗;三、Λ,V,T, θ之 F 考驗;四、 實例分析;五、四種考驗統計量數之抉擇。 |
英文摘要 | There are three reasons why multivariate methods are so important in research. First, multivariate methods limit the inflation of type I experimentwise error rate. Second, multivariate methods honor the nature of reality. Third, multivariate methods take into consideration the intercorrelations among the dependent variables. The four classic multivariate analysis of variance (MANOVA) tests of statistical significance are Wilks' lambda (Λ ), Hotelling trace (T), Pillai's trace (V), and Roy's greatest root (θ )., V, T, and are shown is SAS and SPSS printout for MANOVA. The subjects of this paper including as follows: (1) the role ofeigenvalue in MANOVA; (2) multivariate test statistics; (3) F-test approximation to,V, T, and θ; (4) examples of SAS and SPSS, and (5) choosing a multivariate test statistic. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。