Sorry, I think I sent this message to the wrong address.
James -
I think this is what you're looking for. I actually wrote this syntax in my first year of grad school We were using either SPSS 10 or 11 then, but I know this syntax works on 16.
UNIANOVA
dv BY f3level f2level
/METHOD = SSTYPE(3)
/INTERCEPT = INCLUDE
/PRINT = DESCRIPTIVE ETASQ OPOWER HOMOGENEITY
/CRITERIA = ALPHA(.05)
/lmatrix '3 factor at f2 level 1'
f3level 1 -1 0 f2level*f3level 1 0 -1 0 0 0;
f3level 1 0 -1 f2level*f3level 1 0 0 0 -1 0;
f3level 0 1 -1 f2level*f3level 0 0 1 0 -1 0
/lmatrix '3 factor at f2 level 2'
f3level 1 -1 0 f2level*f3level 0 1 0 -1 0 0;
f3level 1 0 -1 f2level*f3level 0 1 0 0 0 -1;
f3level 0 1 -1 f2level*f3level 0 0 0 1 0 -1
/DESIGN = f3level f2level f3level*f2level .
What you're doing is creating contrasts in your ANOVA. You'll get the results of your anova followed by a series of tables called "Custom Hypothesis tests". I can explain what each line of the syntax is doing, but then I'd be using many generic titles (3factor level 1, 2factor level 2, etc.) and I'm worried that might be confusing. Obviously, you'll need to fill in your variable names but hopefully the labels I've provided make sense to you. If not, shoot me an email, either through the list or back channel, and I'll try to help you out.
Sara House
[hidden email]
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
