Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

+1 vote

This could be done with skips or with a quota question, depending on the exact behavior you are after.

With skips, you would just need to use a skip with logic like this:

That will give each respondent a one-fifth chance of being skipped. But with how probability works, you might observe some significant deviance from the 20% goal going this way.

Quotas can't guarantee that you'll end up with exactly 20%, but they can keep it as close to 20% as possible. What you'll want to do if you go this route is create a quota question with two quota cells. One cell's limit should be four times higher than the other's. Now, the quota question can be set to "Least Fill (Percentages)." Finally, you can add skips that skip respondents based on whether they wound up in the first quota cell or the second.

With skips, you would just need to use a skip with logic like this:

RandNum(RespNum(), 1, 6) = 1

That will give each respondent a one-fifth chance of being skipped. But with how probability works, you might observe some significant deviance from the 20% goal going this way.

Quotas can't guarantee that you'll end up with exactly 20%, but they can keep it as close to 20% as possible. What you'll want to do if you go this route is create a quota question with two quota cells. One cell's limit should be four times higher than the other's. Now, the quota question can be set to "Least Fill (Percentages)." Finally, you can add skips that skip respondents based on whether they wound up in the first quota cell or the second.

...