Re: ANOVA effect size

joanne.williams66@xxxxxxxxx wrote:
I'm currently running a 3-way ANOVA and my sample size is very large
(~16000). Consequently, practically every main effect and interaction
is significant. However, two of the main effects and their interaction
explain the lions share of the variance (the F ratios are F1 = 18, F2
= 53, F1xF2 = 67). All the other effects have F values less than four.
I feel like there should be a short cut way to state this, e.g. "The
main effects and interaction of two of the experimental manipulations
accounted for X% of our explainable variance". Is there such a
measure? Is it as simple as adding up the MStreatments associated with
1, 2, 1 x 2 and dividing by the sum of all the MStreatments? I feel
like that should be a meaningful measure, but perhaps I'm missing
something.


The usual measures of effect size for ANOVA models are eta-squared, partial eta-squared, and omega-squared. You should find hits on those terms with Google, and they will be covered in a lot of stats books for psychology (e.g., Dave Howell's book).

--
Bruce Weaver
bweaver@xxxxxxxxxxxx
www.angelfire.com/wv/bwhomedir
"When all else fails, RTFM."
.