Have an idea?

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

The mean difference between segments for each attribute

Hello all,
I have a question that maybe it’s already discussed, but I’m a bit confused to adjust it to my case of study. I did a CBC with seven attributes with 2-3 levels in each attribute, and then run the HB and LC. I have obtained the average importances for each attribute from the HB, and obtained seven segments from the LC, with raw utilities value for each level in a segment.

Now, I want to know the mean difference between segments for each attribute. For example, the mean difference between segments for attribute A. I'm thinking of running the ANOVA with SPSS (because I think it's more than two segments). Is this correct? If it is correct, what data should I take from the lighthouse studio as an input to SPSS?  

Thank you..

Best regards
asked May 4, 2021 by Adhitya

1 Answer

+1 vote

From LHS export the utility report for the HB analysis.  It has a tab for respondent-level importances.  Those are what you want to import into SPSS to merge with your segment variables (which you can also export from LHS in the LC utility report).  

And yes, with >2 groups ANOVA will be appropriate.
answered May 4, 2021 by Keith Chrzan Platinum Sawtooth Software, Inc. (108,675 points)