Have an idea?

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

Utility/importance of levels across variables varies with counts

Could it be possible that utility/importance of levels across variables varies with counts, what could be the possible reasons...

Only thing I can think of is we estimate utilities on overall basis and not on segment wise/variable wise basis so the variation could be lost when we look at it that ways
asked Jan 18, 2017 by Rajesh Rana Gold (24,000 points)
retagged Jan 18, 2017 by Walter Williams

1 Answer

0 votes
Counts averages across people.  The utilities estimated from logit, latent class, or HB should be fairly similar to what you are seeing from counts (after you exponentiate the utilities...or alternatively after you take the natural log of the counts proportions).  

But, importances from HB are estimated at the individual level.  This can be very different (and more correct) than importances computed from aggregate tables of utilities (across people and segments) or importances estimated from aggregate counts.

If you have an attribute like brand or color where people disagree strongly about their preferences, on average it tends to damp and wash out this attribute to make it appear like it is less important than the individuals and segments behind the averages actually think.
answered Jan 18, 2017 by Bryan Orme Platinum Sawtooth Software, Inc. (169,815 points)