Frequency of combinations possible

I have an SPSS data set with 500+ respondents and 18 symptoms that they could have.

Each symptom has its own variable Symptom01 = 1 means they have the symptom 1 Symptom02 = 0 means they dont have the symptom 2 etc etc

What I want to know is what combination of 3 symptoms is more frequent in my data set. For example how many people have symptom 1, 5 and 6; how many people have symptom 1, 2 and 3, etc.

I doesn't mean that they only have those symptoms. Theey could have others. I just want to know which group of 3 symptoms is more frequent in my dataset.

Can someone help me?


Active Member
Quite likely there isn't any quicker way than just counting them.
However, there are less than 1000 symptom combinations so if you or a friend have basic programming skills it shouldn't take too long to write a short macro with four nested loops using something like vba to run through the lot.
To do it without programming I think you could make 18 17x17 contigency tables (setting a filter every time for x=1, where x is variable n) and then look for the highest number in all of those. Maybe there's an easier way.