PSYC 305 Lecture Notes - Lecture 5: Squared Deviations From The Mean, Analysis Of Variance, Null Hypothesis
Document Summary
The purpose is to test whether the means of k (>2) populations significantly differ: h0 = mu1 = mu2 = muk. We are taking all the variability in our variable and splitting it up into pieces: assess the relative magnitude of the different variance parts. Main goal is to compute f statistic from the given data to make inferences on the null hypothesis. F = msm/msr: mean squares for the model (msm, mean squares for the residual (msr, when need some ss on the top, divided by some df. Overall grand mean (if you ignore the group that people are in, you just compute the mean from all your observations) Group means (3 different groups, compute the mean for each of these groups) Squared deviations between the overall grand mean and the group means. Ssm = sum(ng(xg x)^2: need to know the sample size. Ex: n1 = 7, n2 = 9, n3 = 8 (all nsubg"s) n = 24.