What's the best estimation option?

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

sirooskhadem
Posts: 22
Joined: Tue Nov 19, 2013 12:23 am

What's the best estimation option?

Postby sirooskhadem » Fri Nov 11, 2022 9:07 pm

Hi All,

I need to run a regression where the dependent variable is a rating number (simply 1 to 5) and independent variables are 4 ratios which when we add them all they will always be equal to 1.
Attached is a snapshot of the data and I would be grateful if you could help me picking the right estimation method and also if I need to take account for any statistical/econometric issues?

Cheers,
Siroos
Screenshot_20221112_145218_Chrome.jpg
Screenshot_20221112_145218_Chrome.jpg (459.22 KiB) Viewed 10499 times

startz
Non-normality and collinearity are NOT problems!
Posts: 3779
Joined: Wed Sep 17, 2008 2:25 pm

Re: What's the best estimation option?

Postby startz » Sat Nov 12, 2022 7:32 am

The best option is an ordered logit or probit. Just be sure not to include a constant so as to avoid perfect multicollinearity.

sirooskhadem
Posts: 22
Joined: Tue Nov 19, 2013 12:23 am

Re: What's the best estimation option?

Postby sirooskhadem » Sat Nov 12, 2022 1:37 pm

Thanks very much!


Return to “Econometric Discussions”

Who is online

Users browsing this forum: Bing [Bot] and 1 guest