Re: st: multiple regression, r squared and normality of residuals

[David Greenberg <dg4@nyu.edu>]

Keith, I tried looking for the lnskew command and couldn't find it. Could you indicate where it is located? Thank you, David Greenberg, Sociology Department, New York University 

----- Original Message -----
From: Keith Dear <keithdear4@gmail.com>
Date: Tuesday, March 22, 2011 11:48 pm
Subject: Re: st: multiple regression, r squared and normality of residuals
To: statalist@hsphsun2.harvard.edu


> You could also try a sqrt transform, since the log seems to have
> overcooked it: see -ladder-. And you appear to have seven zeros (or
> negatives) which is why you are losing N: a common solution is to use
> log(1+x); or try -lnskew-
> 
> However since your original residuals appeared normal, why are you
> transforming the dependent variable at all? You probably have
> something other than a straight-line dependence on one or more
> covariates, so should be concentrating on the RHS of the equation not
> the left. -fracpoly- and -mfp- may help, or or just lots of
> scatterplots. Also see -help regress postestimation- for -avplot- and
> others.
> 
> Keith
> 
> On 23 March 2011 12:11, Arti Pandey <rtpandey@yahoo.com> wrote:
> >
> > Hello
> >
> > I ran multiple regression with in stata using two models;
> > the first gave an R-squared of .35, p values of all predictors was 
> less than
> > 0.001 except one which was less than 0.05. No.  of obs. used was 84,
> > distribution of residuals was normal.
> > Then I did a log transform of the dependent variable, r squared went 
> up to .65,
> > p values for all predictors was 0.001 except the one mentioned 
> above, which is
> > now 0.06. The residuals were also slightly skewed to the left. No. 
> of obs went
> > down to 77.
> > My question is how do I decide between the R squared and 
> distribution of
> > residuals. Is such a high rise in R squared worth sacrificing no of 
> observations
> >
> > and normal distribution of residuals for. Since the skew is not very 
> pronounced,
> >
> > it is tempting to go with the second, but then the regression  model 
> might be
> > wrong.....
> > Appreciate any help.
> > Arti
> >
> >
> >
> >
> > *
> > *   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/
> >
> 
> 
> 
> -- 
> Dr Keith Dear
> National Centre for Epidemiology and Population Health
> ANU College of Medicine, Biology and Environment
> Australian National University
> Canberra, ACT 0200 Australia
> CRICOS provider #00120C
> Phone +61 (02) 6273 2208
> Mobile 0424 450 396
> 
> *
> *   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/

*
*   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/