Have an idea?

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

Running correlation using ZCdiff individual utility values

Dear CA experts,

I plan to run correlation analysis for several IDV (all measured using 7 point likert scale) with individual utility values, & possibly run MRA and MANOVA accordingly.

Upon reading the technical papers, I should use Normalized (zcdiff) utility values to capture the heterogeneity of my samples, correct?

Would it be the right approach to run correlation analysis for each level in attribute? Since averaging the utility value per attribute will all come to "0"?

My scenario has 5 attributes and a total of 13 levels.

Thanks for any kind of input.
asked Jun 28 by anonymous

1 Answer

0 votes
Depending on what your research objectives are, you could run correlations on the levels' utilities or on the attribute importances.  And yes, I would use the ZCD utilities for running correlations.
answered Jun 29 by Keith Chrzan Platinum Sawtooth Software, Inc. (95,775 points)
Thank you for your reply!!!! Have a good day.