Hi everyone.
I have the following problem. In an experiment I had a number of plants of two developmental stages. I applied a treatment with two levels, so there are 2 independent variables (dev. stage and treatment) with 2 levels each, which makes 4 groups of plants. On 3 different sampling dates I sampled randomly a specific number of plants. So “Sampling date” is the 3rd independent variable with 3 levels. I used it to examine the impact of time, although I used different plants on each sampling due to destructive sampling.
I use SPSS. To test the impact of dev. stage, treatment and sampling date I ran a GLM analysis with 3 fixed factors. I was advised to use individual as “random factor” to avoid temporal autocorrelation problems, but I did not, because the plants of each sampling date were not the same ones. I would have used GLM Repeated Measures if that was the case. Is that correct? But if I had to use individual as random factor, would I have to add something on the sav. sheet?

Thank you!