SPSS Statistics

Hello, 

Do you have any ideas what kind of statistical test I can use in SPSS Statistics to compare three proportions with partially overlapping samples:

I have two types of variables:

A. Target market (nominal variable)

1. Students
2. University personnel
3. Both (mixed segment)

B. Marketing mix (measured through twenty dichotomous variables + two ordinal variables)

I want to measure if there are differences between the three target markets regarding the marketing mixes used by companies. I initially applied the Chi-square test of homogeneity & Post hoc test of z-values comparing the proportions, however,  I. later realized that the three target markets were not independent and I did not have the right to use this statistical test. Therefore, I was wondering if someone can give me any ideas which test and post-hoc test might be relevant in my case ad are available in SPSS. The main problem is basically that the third category "Both" includes the first two ones, therefore the the three groups are related even though not all of the three groups come from the same population. 

Thank you in advance for your reply,