Fixed and Random Effects in Mixed Modeling

#1
Thank ahead of time for any help.

I am analyzing time series data. Originally I was running Repeated Measures ANOVA and entering between subjects variables. This was providing significant and interesting results. However, there is missing data, and of course, RM-ANOVA deletes listwise. I was digging around and discoverd mixed model ANOVAs. I've spent two days reading and learning about mixed models and I think they may be benifitial to my situation. However, I am not confident in how I am entering variables.

Participants filled out personality questionnaire (I'll call this "BI").
Participants were then split into squads of 4-8 members (Squad).
They were then randomly assigned as a leader or a follower (Leader).
I have cortisol levels at 4 time points (Cort).

Ultimately, I want to investigate differences in cortisol responses across time. I believe that BI will impact the strength of the response as well as leader. Members of the same squad experienced the exact same exercise, while those in different squads experienced the same exercise, but at a different time.

I am using SPSS

I don't know how to set the model up. Here is what I have tried:

First I restructed my data to long form.

-Analyze / Mixed Models / Linear.
-The first box that comes up I enter "ID" into the subjects box and "TIME" into the repeated box.
- The second box I enter Cortisol as the Dependent variable. I then enter SQUAD, Leader, and BI into the factors box.

Should squad be a covariate?
Which variables are fixed factors and which are random?


Thanks again,
John