Have an idea?

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

Affects of covariates on attributes (ACBC)

Thanks again so much for generous help provided! I am truly grateful!

I used an ACBC survey (without BYO) and I included few moderators. I want to see how they are affecting the attribute choices. I wanted ask for an advise, if possible, about the approach that needs to be taken to find these affects.  
I was thinking to compare the HB estimations with and without covariates and report the findings based on the differences of these two output. Is there a more relevant way to do that?

Thank you !
asked Dec 10, 2014 by Mir (400 points)

1 Answer

0 votes
If you want to use Bayesian statistics to investigate the effect of covariates on utilities, then review the paper: http://www.sawtoothsoftware.com/downloadPDF.php?file=HBCovariates.pdf

If you want to use Frequentist statistics to do this, then you should use the zero-centered diffs normalized utilities.  You'll find that our software outputs point estimates (single accumulated set of scores per respondent) for the zero-centered diffs.  Use zero-centered diffs, because it is an attempt to get rid of the large scale factor differences that can be observed across respondents.  Then, read the zero-centered diffs utilities along with the segment membership info into a standard cross-tab package (or statistical package) and conduct F-tests, T-tests, or ANOVAs between segments.
answered Dec 10, 2014 by Bryan Orme Platinum Sawtooth Software, Inc. (198,415 points)
Affects of Covariates and cryptic error