Outlier detection

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

Weaver102406
Posts: 13
Joined: Tue Aug 30, 2011 10:10 am

Outlier detection

Postby Weaver102406 » Tue Aug 30, 2011 10:16 am

I am attempting to detect outliers for my data, but I do not see the 'Stability Diagnostics' option when I click on View. What do I need to do? Also, once I am able to execute the leverage plots and influence statistics, what metric do I use to determine an observation is an outlier?

Thanks,
Carol

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13604
Joined: Tue Sep 16, 2008 5:38 pm

Re: Outlier detection

Postby EViews Gareth » Tue Aug 30, 2011 10:24 am

Stability Diagnostics is only available for equations (and only for certain estimation methods), and in EViews 7. If you don't see it, and you're using EViews 7, it is probably the case that you're not looking at an equation object.

Weaver102406
Posts: 13
Joined: Tue Aug 30, 2011 10:10 am

Re: Outlier detection

Postby Weaver102406 » Tue Aug 30, 2011 10:32 am

I am using Eviews 7 and estimated a least squares equation..both ordinary and with random effects, but do not find the Stability Diagnostics option. I have used it in the past and am not sure what I am doing wrong today. :)

Thanks!

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13604
Joined: Tue Sep 16, 2008 5:38 pm

Re: Outlier detection

Postby EViews Gareth » Tue Aug 30, 2011 10:34 am

It isn't available for panels.

Weaver102406
Posts: 13
Joined: Tue Aug 30, 2011 10:10 am

Re: Outlier detection

Postby Weaver102406 » Tue Aug 30, 2011 10:37 am

What method would you suggest for detecting outliers in panel data since these are not available?

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: Outlier detection

Postby EViews Glenn » Tue Aug 30, 2011 10:54 pm

Part of the reason it is not supported is that once you get to panel data the concept of an outlier becomes more vague. We are not particularly comfortable advocating an outlier method in this case.

Weaver102406
Posts: 13
Joined: Tue Aug 30, 2011 10:10 am

Re: Outlier detection

Postby Weaver102406 » Wed Aug 31, 2011 8:30 am

Thanks...I'll check with my professor for his recommendations.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests