Chapter 13 Analysis of Variance and Experimental Design

  • Slides: 41
Download presentation
Chapter 13 Analysis of Variance and Experimental Design

Chapter 13 Analysis of Variance and Experimental Design

Assumptions for Analysis of Variance • For each population, the response variable is normally

Assumptions for Analysis of Variance • For each population, the response variable is normally distributed. (應變 數都要是常態分配) • The variance of the response variable 2 , denoted σ , is the same for all of the populations. (所有母體應變數之變異 2 數 σ 都要相等

利用MSTR/MSE的抽樣分配計算p值 MSTR/MSE的抽樣分配 p值 F=9. 00 MSTR/MSE

利用MSTR/MSE的抽樣分配計算p值 MSTR/MSE的抽樣分配 p值 F=9. 00 MSTR/MSE

ANOVA表 表 13. 2 H公司例子之變異數分析表 Source of Variation Sum of Squares Degrees of Freedom

ANOVA表 表 13. 2 H公司例子之變異數分析表 Source of Variation Sum of Squares Degrees of Freedom Mean Square F Treatments 516 2 258. 00 9. 00 Error 430 15 28. 67 Total 946 17

Fisher’s Least Significant Difference • 費雪LSD程序: H 0: μi＝μj Ha: μi≠μj • 檢定統計量 t=

Fisher’s Least Significant Difference • 費雪LSD程序: H 0: μi＝μj Ha: μi≠μj • 檢定統計量 t= x 1 - x 2 MSE( 1 + 1 ) ni nj