I have tried to look for the best way to do this with no success so I thought I would ask before going slightly mad ....

What I need to do is randomise groups and subgroups within the initial groups with all respondents seeing all questions ..

So we have a structure as follows

Product	Feature	Question
P1	F1	P1F1Q1
		P1F1Q2
		P1F1Q3
	F2	P1F2Q1
		P1F2Q2
		P1F2Q3
P2	F1	P2F1Q1
		P2F1Q2
		P2F1Q3
	F2	P2F2Q1
		P2F2Q2
		P2F2Q3
P3	F1	P3F1Q1
		P3F1Q2
		P3F1Q3
	F2	P3F2Q1
		P3F2Q2
		P3F2Q3

So a respondent would see randomly a Product Then Randomly a Feature but then answer the questions in order.... and then see the other feature and then the other products...

Outside of developing a survey for each of the 3 products and then randomising the Feature Groups I cannot seem to find a way of doing this easily ...

Is there some clever or interesting way to accomplish this that I am missing?

Thanks in advance.

Unfortunately, there is currently no way to randomize "sets" of groups.

Thanks Tony ...

In the meantime I think I have come up with a solution that works (loose definition of works) for me.

What I have done, and the people processing this post collection will hate me for it, is as follows....

1) Set up a Multiple Short Text Question with 3 options
2) Use an equation question to generate a random number
3) Use this random number to populate the short text answers
4) Randomise the sub groups as mentioned using the group randomisation option
5) Use the text replacement to feed in the populated answer from step 3
6) Repeat for each sub group and each of the 3 products.

This seems to have the desired effect for data collection and will just need the data processor to post process which of the products was seen 1st 2nd or 3rd and align the data accordingly.