using utilities of different attribute levels (from conjoint analysis) in regression?

#1
In short, I am trying to measure factors influencing an intention-to-purchase a product, with use of a regression. However, some of my constructs were measured with conjoint analysis and resulted in a couple of variables with utility values for all of them. What I mean is that e.g. if I have given them a choice between energy portfolios having 5%, 25%,50%, 75%, 95% of renewable energy - now, instead of ending up with one choice (and one variable), I have respondent's utilities for each of those levels. Thus, I was wondering how do I implement it in my regression analysis: do I treat them as separate independent variables and calculated betas for each? Somehow it does not feel right. But then again, how do I "combine" them into one variable?

Thanks in advance for any advice!

Basia
 

Masteras

TS Contributor
#2
Re: using utilities of different attribute levels (from conjoint analysis) in regress

So one independent variable had values 5%, 25%, 50%, 75% and 95% only? 5 distinct values only?