Hi,
I'm currently running an ACBC study with 6 Product attributes, and one price attribute. Each of my six product attributes are "include or don't include" attributes, so essentially respondents could create a bundle in the build your own that had no products in it. Many did. Many also continuously said all the bundles that had anything in them in the screening questions would not work for them.
My questions are about the treatment of the none parameter
1. What is the full interpretation of the None parameter?
2. I have the ability in simulations to compare any product bundle to the "bundle" that does not have any products in it and thus costs zero dollars. What are the implications if I were to take out the none parameter and simply compare bundles against the empty "bundle" that costs zero dollars? How is this different than comparing the share of preference for one bundle in a simulation versus the percent that would choose "none of the above" that is estimated with the None parameter?