Have an idea?

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

Dual response none

Hi community,

I am conducting a CBC with a dual response none.
In the first stage of my CBC (forced choice) I am have 3 attributes with 4 levels each. Every alternative in the first stage is representing a reform option of a health law.
In the second stage (the dual response none) I am asking whether the respondents would really vote for the reform option they have choosen in stage 1.

Is that a reasonable approach? Having in mind that not voting for the reform option indirectly means staying with the current version (status quo) of the law and two of the attribute levels that are defining the status quo are also part of the CBC alternatives of stage 1.
I am not sure in how far those attribute levels would be estimated double or harm the estimation?

Looking forward to receiving your advice.
Thank you in advance.

asked May 3 by anonymous

1 Answer

0 votes
Best answer

I think this makes plenty of sense.  The none option in conjoint analysis, be it dual response or not, is often a status quo option.  That the status quo shares some elements of the experimentally designed profiles is also common and not problematic that I'm aware of.
answered May 3 by Keith Chrzan Platinum Sawtooth Software, Inc. (102,700 points)
Dear Keith,

thank you very much for your quick and helpful response.

So, also with regards to analysis I could interpret the pooled estimates for the attribute levels of the 1. stage alternatives without considering that some of the levels are part of the status quo as well? And I suppose that I could also compare the none coefficient to other reform alternatives in the market simulator, couldnĀ“t I?

Thanks again for your help.
Yes, the status quo has it's own utility, captured in the alternative-specific constant that represents the None alternative.