APMA 3120 Chapter Notes - Chapter 15: Dependent And Independent Variables, F-Distribution, Test Statistic

• School: University of Virginia
• Department: Applied Mathematics
• Course: APMA 3120
• Professor: James Lark

Document Summary

Populations are assumed to be normal equal standard deviations 1 2= 1 = 2 = . [rule of thumb: check that largest sample standard deviation si is no more than twice the smallest], but possibly different means. each group mean: measure variation among group means from overall mean & variation within groups from. Of squared deviationsamong groups(ssg)= of a group mean group, then takes the overall sum from the overall mean. I ni( xi x)2 weights each squared deviation i=1. X with the number of observations ni for that. Mean of squared deviationsamonggroups (msg ): dfg=i 1 msg= ssg. Mean sum of squared error within groups (mse): sse has degrees of freedom dfe = n-1 so. Dfe: f statistic ratio is used to test e, h0: population means for all groups are equal [ 1= 2= 3= i, ha: not all the i are equal. F= mse close to 1, indicating variation among groups is not significantly greater than.