Hi David,

I posted a quick response to your previous solution where we could use Distict counts of respondents, recall if a respondent responds more than once to a combination we count them only once. Your previous solution with the small amendment works quite well but does not solve the problem where we have 50 + products, got any ideas?

quote:

---

Having trouble with the logic here.

If respondent has A & B then you expect

A 1

If respondent has A & C then you expect

A 1

If respondent has B & C then you expect

B 1

If so you will never get AB combination

Sorry if I seem thick here but I am trying to get a handle on this.

---