PSYC3301 Lecture Notes - Lecture 2: Factorial Experiment, Null Hypothesis, Sphericity
Document Summary
Anova is a f ration- the ratio of the good stuff to the bad stuff. To find the variability among the group means use the sum of squares. Sum of squares within= ssw= sum of (nj-1) sj^2. Degrees of freedom- ssw and ssb were calculated on different scales. Thus must divide them by their subsequent degrees of freedom to make them equal. All f values are positive, cannot be negative. Eta squared (n^2) measure of effect size for anova one way. N^2 represents the proportion of variance in the dv that is accounted for by the iv. Omega squared is w^2- another measure of effect size- not bias. Between groups anova- more than 3 groups and multiple comparison. Bonferroni correction for anova- divide 0. 05 by the number of comparisons then see if still significant. Fishers lsd- only handle 3 means- equal variance. Planned comparisons- are hypothesis driven- need past empirical research.
