頁籤選單縮合
題 名 | On Assessing the Assumption of Normality=談常態性假設之評估 |
---|---|
作 者 | 陳榮治; | 書刊名 | 嘉義技術學院學報 |
卷 期 | 64 1999.06[民88.06] |
頁 次 | 頁201-212 |
分類號 | 319.5 |
關鍵詞 | 常態分配; 母體; 樣本; 中央極限定理; 機率密度函數; 平均數向量; 共變數矩陣; 相關係數; Q-Q圖示法; 卡方圖示法; Normal distribution; Population; Sample; Central limit theorem; Probability density function; Mean vector; Variance-covariance matrix; Correlation coefficient; Q-Q plot; Chi-square plot; |
語 文 | 英文(English) |
中文摘要 | 常態分配在統計學上是最重要且被廣泛使用的分配,因為它常被考慮為許多類不 同的量度之母體模式,而且,很多重要的程序性質,均建立在母體想像成至少接近於常態分 配這個假設基礎條件上,因此,尋找一個較正式的程序,來檢查樣本資料是否來自於常態分 配, 也就顯得頗為需要。 此處所要介紹的檢驗方法, 是非常實用, 易於使用統計軟體 MINITAB 在 PC 上來執行,而且,也足以勝任非常態性的偵測。 本文中,我介紹了一些檢驗常態性假設的技巧,尤其,基於實用上考慮,重點擺在檢驗單變 量常態分配 -- 用 Q-Q 圖示法,及檢驗雙變量常態分配 -- 用卡方圖示法, 並且,用一個 實例來說明其執行程序。最後,並分別作出對應圖形以便觀察與判斷。 |
英文摘要 | The normal distribution is the most important and widely used distribution in statistics because it is often used as a model for many sorts of measurement, indeed, many of the procedures to be presented depend on the population being at least approximately normal. Therefore, a procedure is really needed for checking that the sample data could have come from a normal distribution. Statistics has some procedures for doing this, but the ones presented here are easy to use and do a good job of detecting nonnormality. In this paper, I principally present some useful techniques for checking the assumption of normality, especially for most practical consideration, I will concentrate on the univariate normal distribution -approach in Q-Q plot and the bivariate normal distribution-approach in chi-square plot. |
本系統中英文摘要資訊取自各篇刊載內容。