STATS 250 Lecture Notes - Lecture 13: Dependent And Independent Variables, Lincoln Near-Earth Asteroid Research, Scatter Plot
49
STATS 250 Full Course Notes
Verified Note
49 documents
Document Summary
We will only do the pooled version of anova. The hypothesis we"ll be testing today is between (cid:885) different drugs. Ho is always about no difference, no effect and thus all the averages are equal. Ha will be that at least one population mean is different. The new statistic is called an f statistic that estimates the common variance. Mse is always a good estimate from your data regardless if ho or ha is true. Ms groups is only good if ho is true so it is a good estimate when ho is true. The ratio will be 1 if ho is true. If ho is not true, ms group will be larger and thus f will be much greater than 1. Compared to the t-statistic which had a t distribution, the f-distribution follow the f distribution. The f distribution is not centered at 0.
