PSYC 4131 Study Guide - Midterm Guide: Multivariate Analysis Of Variance, Observational Error, Analysis Of Variance
Document Summary
Same thing as aski(cid:374)g ho(cid:449) (cid:373)a(cid:374)(cid:455) (cid:272)ells (cid:449)ill the(cid:396)e (cid:271)e. to do this, it"s as si(cid:373)ple as (cid:373)ultipl(cid:455)i(cid:374)g (cid:1006)(cid:454)(cid:1006)= (cid:1008). Omnibus manova, then if it is significant univariate anova, then post- hoc if needed (the type of post hoc also depends of how many levels. If there is 2- t-test, if there is (cid:1007) l d o(cid:396) bo(cid:374)fe(cid:396)(cid:396)o(cid:374)i, if (cid:455)ou get (cid:395)uiet a fe(cid:449) the(cid:374) (cid:455)ou a(cid:396)e goi(cid:374)g to (cid:449)a(cid:374)t to (cid:396)u(cid:374) a tuke(cid:455)"s. For anova or manova know the null vs the alternative hypothesis. I think it is more complicated than that, will have to look in notes. Pair-wise comparisons i think are another word for post hoc, could be asked like that also. If anova is significant there is a main effect difference somewhere. Know when post- hoc tests need to be ran: describe the logic behind an anova. Because there is more than 2 levels being compared we need to control for experiment wise error rate.
