Need Help with GLMM in SPSS

[losibk2]

Hi,
i`m new to SPSS and basically a novice in statistics too. I would be thankful for every kind of help.

I hava a Panelset with 8 Waves. My research question is to analyze the influence of socio-demografic factors on the credit situation. Every subject has an id.

I want to analyse two dependant variables. 1. Credit (binary yes or no) , 2.Creditvolume (metric)

My independant variables are gender, age group, ethnicity, Income groups, schooling and a few more.

My goal is to analyse hypothesis like for example:

1. Males have a higher probability of having credit debt.
2. People with higher income have a higher probability of credit debt.
3. People with children have a higher probability of credit debt than people without children.

I read a lot of literature about random effect/fixed effect models and GLMM in Spss but I wasn't able to understand this topic.

I have read that fixed effect models use the within estimator and measures the differences within a subject and can't be used for time invariant factors like gender or race. In addition fixed effect models work over time. So Im not able to compare for difference in schooling before and only for schooling which was passed during the panel data.
Random effect model use between estimators and measures differences between subjects.

So in my opinion I should stick with a random effect model. Maybe a random intercept model? Is this correct?

I tried to model in SPSS with GLMM. In the first step I insert ID in Subjects and time for repeated measures.
In the next step Im not really able to understand how the fixed and random effects relate to the fixed/random effect models.

If I want to estimate if the credit volume of men is significantly higher than of woman, is the following syntax correct?

MIXED VolCredit BY Gender
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=Geschlecht | SSTYPE(3)
  /METHOD=REML
  /PRINT=SOLUTION
  /RANDOM=INTERCEPT | SUBJECT(respid) COVTYPE(UN)
  /REPEATED=time | SUBJECT(respid) COVTYPE(UN).

Im really thankful for any kind of hints

Regards Markus