Partial Utilities and Linear Regression in Conjoint Analysis

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

Posts: 1
Joined: Thu Dec 06, 2012 5:24 pm

Partial Utilities and Linear Regression in Conjoint Analysis

Postby mapepe » Thu Dec 06, 2012 5:46 pm


I am trying to teach myself how to do a conjoint analysis in SPSS. Currently, I have two problems with it:

1. When I calculate the partial utility for one factor (e.g. the price) and one individual by hand, I get different results than SPSS does. The formula tells me to substract the mean of the ranks for one value from the mean of all values. This would give me the partial utility for that value. In SPSS I get it using the DISCRETE model.
However, SPSS gets a different value and apparently uses a different formula/algorithm. Can anyone tell me how this is calculated in SPSS?
One thing I suspect is that I use an orthogonal plan, i.e. I don't have results for all possible combinations of values. Maybe SPSS somehow interpolates the missing ranks?

2. Once I have these partial utilities (via the SPSS DISCRETE model) I want to calculate the linear regression for them. To my understanding, this is what SPSS does when I use the LINEAR model. Is that correct? If so, why do I get different results again? I take the values the SPSS DISCRETE model gives me and solve the least squares problem.
Does SPSS enhance the results afterwards, e.g. by taking other factors into account?

I would really appreciate if someone could help me with this. I am new to econometrics but I am studying mathematics so I guess you can go into detail here.

Thanks a million!

Who is online

Users browsing this forum: No registered users and 2 guests