Have an idea?

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

Too many prohibitions - already fielded

I didn't realize the adverse affect that prohibitions had on the validity of the conjoint, so i fielded the survey and now the standard errors in the test design are starred out

i know this is bad, but is there any way to salvage the data??
asked Nov 8, 2019 by anonymous

2 Answers

0 votes
Yes, this bad - usually this means you have a disaster on your hands.  

On very rare occasions, however, the situation is salvageable.  If you want to send me your study (keith@sawtoothsoftware.com) I'll be happy to take a look at it.  I'll need the .ssi file, the .cgi file and the .db3 file.

I don't want to over-promise, however, as by far the most likely result is that you'll have a disaster on your hands and I'll advise you to refield the survey.
answered Nov 8, 2019 by Keith Chrzan Platinum Sawtooth Software, Inc. (92,075 points)
0 votes
It's often possible to salvage the data, but it depends exactly on the nature of the  problem. Typically, since the task was logical to the respondents, it's a matter of respecifying the underlying statistical model and rerunning after that's been fixed.

I would recommend talking to our support team at support@sawtoothsoftware.com. They'll need to look at your study files and your data to see exactly what would need to be done.
answered Nov 8, 2019 by Kenneth Fairchild Bronze Sawtooth Software, Inc. (3,820 points)
Sometimes by collapsing two attributes that had too many prohibitions into a single attribute listing the shown two-way combinations of levels (this requires some data manipulation) the data might be able to be saved.  Keith and the support team here at Sawtooth Software are great resources.