Have an idea?

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

Generating Data

Dear team,

I want to use the Data Generator to create random responders to check for consistency cutoffs. Is there any possibility to also include a none option in the random generated data or is this not necessary? I am asking because my original CBC study has a none option included.

Thanks for your help!
asked Aug 18 by anonymous

1 Answer

0 votes
You should make your CBC study for random data generation look just like your real study.  So, if you included (for example) a None or a Dual-Response None, then you should do the same for your parallel study that you use to establish RLH cutoffs to identify near-random responders.
answered Aug 18 by Bryan Orme Platinum Sawtooth Software, Inc. (190,165 points)
Yes, I linked my real study URL to the generating data field, so it's exactly the same for the random data generation as my real study - but when I analyze the random data I can see that None has never been chosen - so I am wondering if I missed any setting in the random data generation? Or what am I doing wrong?
I'd think the random data generator should answer the Nones randomly as well.  Perhaps you can call or write our tech support help line and they'll have some recommendations.