Have an idea?

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

post regression in R, SPSS

I have calculated a CBC HB analysis and want to perform regression analysis with other variablen in my survey. Usually I export the zero centered utilities to R, SPSS merge them with the survey and perform statistical analysis.

So, my dependent variables are the attribute level utilities and my independent variables are those from the survey itself.

However, there seem to be a problem with my results. When i perform linear regression analysis I only get very very small R2 (and adjusted R2) values from variables that "must" perform better as they are obvious to explain more variance then represented by my result.

for example: one attribute level is a "landscape type 1" (dependent) and the independent variables come from variables about meanings people assign to landscapes (to the same landscape type as the dependent variable). So I know that those variables are important for the decision of which landscape is chosen, but the result shows me that it is not.

Know I am thinking about why this could happen?

I really tested a lot of variables (also factors calculated out of item boxes and items that are referenced in literature) but I am stuck.

Am I using the false dependent variables?  Shouldn't  I use linear regression? Or is there any other tipp/trick you can provide me to proceed further?

Thanks a lot for your support.
asked Mar 14 by bs77 Bronze (710 points)

Your solution to the original question

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.