Greetings to any SPSS geniuses out there in ecoland:

I’m currently working my MS thesis, and need help on how to proceed with my 
analysis in SPSS:

I took samples once a month for four months. I have two locations with the 
same experimental design. Each location has four treatments (control, low, 
med, high), not randomly assigned. Each of the four treatments is made of 
three plots. Samples were taken from 3 randomly placed quadrats within each 
plot.

Here’s what I think is going on:

three factor mixed model ANOVA

Month is a factor (fixed, not repeated measure!) Location is a factor 
(fixed) Treatment is a factor nested in location (random) Quadrats are my 
replicates Like any eco-data, I have unequal sample sizes (not all 
replicates were captured every month)

y = month + location + treatment(location) + month*location + 
month*treatment(location) + location*treatment(location) + 
month*location*treatment(location)


1. Should I use the general linear model or the mixed linear model? a. GLM 
won’t let me use type III SS, lets me use type I SS. Then I found via 
google that GLM isn’t best for nested designs, a mixed linear model. But I 
have **NO IDEA** how to enter everything in to the mixed model dialog 
boxes. PLEASE HELP! b. What about type IV SS?
2. How do I enter all of the interaction terms when using a nested model?


Please have pity on me, as I’ve already spent too much time trying to 
figure this out on my own!!! Sincerely, Pamela Archer

Reply via email to