Have an idea?

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

Low average and median r squared


Just ran OLS for CVA questions and got the results of .44 average r squared and .40 median r squared.

Is this indicating that CVA scenarios are designed poorly and pose a big concern?

Possibly my attributes were not comprehensive?

Thank for the reply.
asked Jun 28 by anonymous

1 Answer

0 votes
The r-squared speaks directly to the internal consistency of respondents in answering the conjoint profiles.  It is mostly independent of how comprehensive or not your attributes were regarding coverage of the aspects involved in making choices for your product situation.  Though, a less comprehensive conjoint attribute list would likely lead to higher internal r-squared, because it would be easier for respondents to be internally consistent if they had to worry about fewer attributes varying at the same time in the experimental design.
answered Jun 29 by Bryan Orme Platinum Sawtooth Software, Inc. (177,015 points)
Thank you so much!!!!