Hello,
In my survey, I have a matrix question with 6 prompts. These are all framed in a positive way and look something like this
how much do you agree with the following statement?
1) "This policy would be good for the environment."
2) "This policy would be good for business."
3) "This policy would be good for employees."
4) "this policy would be good for investors."
etc..

I would like to present to respondents this matrix with 3 levels of randomization. The third and the are the ones I am struggling with.
a) They should only see a random set of 3 (ot of 6) prompts
b) These should be presented in random order
c) They should vary the positive/negative structure of the sentence in a random way. (e.g. This policy would not be good for the environment.)
d) No prompt (regardless if it is positive or negative) should appear twice.

Hence respondents see something like this: where the order, the content, and the positive/negative framing is randomly allocated.
1) "This policy would not be good for business."
2) "This policy would be good for investors."
3) "This policy would not be good for employees."

Does anyone know how I could do this?