EViews.com

Dear all,

I am wondering whether and how I can integrate brand specific random effects in my regression with cross- sectional data.
My dataset consists of 340 observations (y) that belong to 70 brands. Observations per brand range from 1 to 20. My main goal is to explain y by a number of independent variables (x). In order to control for brand specific effects I like to including a random error per brand.
Thus, my question is how can I incorporate random errors for cross-sectional data in eviews?
Your help would be highly appreciated.
Thank you
Maren

You'd have to structure your workfile as a panel, where the "date" is an Interger Date running from 1 to 20, and your brands are the cross-sections.