Have an idea?

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

Multi-level modeling: Keeping one attribute constant

Dear all,

we are currently working on the analysis of a forced choice based conjoint, where participants were shown two companies that differed in seven attributes and they had to choose which one they would prefer. Now, we have the following idea and I would like to ask you, if this could cause some trouble or if it can be done this way:

We are using multivel regressions with clustered standard errors on each individual in Stata, to estimate log-odds for each attribute. In our next step, we want to keep one of the attributes constant, i.e., business model. There were four different business models, and we want to estimate the log-odds for cases, where for example only a low-cost business model is shown and then compare the estimates for each of the models.

I am currently a little unsure, if it can be done this way. Actually, it is someway a "test for interactions" between presented attributes, as far as I know, this would require simulation based modelling.
Nevertheless, could someone tell me if the results from these models can be used and compared? (It is of special interest for us, if the importance of certain attributes change, if the business model is a different one).

I would be very happy for any comments!

asked Jan 21 by wd (120 points)

Your answer

Please only use this to answer the original question. Otherwise please use comments.
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
Anti-spam verification:

To avoid this verification in future, please log in or register.