Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Maarten Buis <maartenlbuis@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Fixed effects regression and r-squared values |
Date | Mon, 10 Oct 2011 16:44:46 +0200 |
On Mon, Oct 10, 2011 at 4:15 PM, <martin.r.carey@uk.pwc.com> wrote: > Has anyone an idea why a fixed effects regression with 'outliers' included > would have an R2 of 0.025 and then with 'outliers' removed and it jumps to > 0.940? R2 is the proportion explained variance. If you exclude `outliers' you will (greatly) reduce the total variance in y, and since you labelled them as outliers, the removed variance was in all likelihood not well explained by your model. In those cases you would expect the R2 to increase. However, in general you should only compare R2s when the models are estimated on exactly the same sample _and_ you use exactly the same dependent/explained/left-hand-side/y variable. Moreover, even though I would expect an increase, that change looks suspiciously large to me. Hope this helps, Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/