Lec6_chap12_ANOVA英文电子资料.pptVIP

Chapter 12: One- Way Independent ANOVA Face to multiple groups…… 运动治疗 针灸疗法 ........... 西药 Hypotheses (omnibus test) H0: u1 = u2 = u3 Ha: u1 ≠u2 ≠ u3 H0: u1 = u2 =….. Uk Ha:not all us are equal As with the t value from a two group test, the F ratio gets larger as the separation of the group means gets larger relative to the variability within groups. Before we can make any judgment about when F is considered large, we need to see what happens to the F ratio when Dr. Null is right or the treatments have no effect at all. F distribution is really a family of distribution, as ratio of two distributions. F distribution changes shape depending on the number of groups, as well as the total number of subjects. The one-tail F test is equivalent to a two-tailed t test at the same alpha (when k=2) Regardless of which mean is larger, a big difference between the two means will lead to a larger positive F Sample Estimates Population  MSw  MSb + Relative spread of three population distributions Advantage of the one-way ANOVA The chance of making a type 1 error increases as more pairs of means are tested. ANOVA can find a significant difference among several group means even when no two of the means are significantly different from each other. Steps of ANOVA Step1. State the hypothesis Testing homogeneity of variance two group Multigroup Effect size in ANOVA Varieties of the one -way ANOVA Fixed vs. Random effects IV是特别挑选，感兴趣的水平 IV是随机挑选（药剂量） Qualitative vs. quantitative levels 当因变量和因变量都是数量变量时,方差分析是回归分析的特例. 此时,回归分析更有效度. Why is regression usually more powerful than ANOVA with quantitative IV levels 当所有的组均数都落在一条直线上的时候，SS regression