PSYC3010 Lecture Notes - Lecture 1: Analysis Of Variance, Factorial Experiment
Document Summary
Factorial designs: combine two one-way experiments, has at least 2 factors (ivs) each with at least 2 levels, two ivs can be examined simultaneously. One-way anova: one iv, group means compared to each other & to the grand mean. If different there is an effect of the iv. The grand mean: the mean of all observations and forms part of the structural model of anova. Marginal means: a significant main effect of example type tells us marginal means are different: follo(cid:449) up (cid:272)o(cid:373)pa(cid:396)iso(cid:374)s a(cid:396)e (cid:374)e(cid:272)essa(cid:396)y as it does(cid:374)"t tell us (cid:449)he(cid:396)e the diffe(cid:396)e(cid:374)(cid:272)es lie. 2 main effects with no interaction (lines are parallel) Then, the main effects may need to be reinterpreted interactions may be ordinal (cid:894)li(cid:374)es do(cid:374)"t (cid:272)(cid:396)oss, (cid:449)ithi(cid:374) the (cid:396)a(cid:374)ge of data (cid:373)ea(cid:374)i(cid:374)gful to the study) or disordinal (lines do cross within that range) Factorial anova: one iv, group means compared to each other & to the grand mean.
