[R] Deviance Residuals

Iasonas Lamprianou lamprianou at yahoo.com
Fri Aug 20 14:23:48 CEST 2010


that's enough, thank you
now I need to do some more background reading... cheers

Dr. Iasonas Lamprianou


Assistant Professor (Educational Research and Evaluation)
Department of Education Sciences
European University-Cyprus
P.O. Box 22006
1516 Nicosia
Cyprus 
Tel.: +357-22-713178
Fax: +357-22-590539


Honorary Research Fellow
Department of Education
The University of Manchester
Oxford Road, Manchester M13 9PL, UK
Tel. 0044  161 275 3485
iasonas.lamprianou at manchester.ac.uk


--- On Fri, 20/8/10, John Fox <jfox at mcmaster.ca> wrote:

> From: John Fox <jfox at mcmaster.ca>
> Subject: RE: [R] Deviance Residuals
> To: "'Iasonas Lamprianou'" <lamprianou at yahoo.com>
> Cc: r-help at r-project.org
> Date: Friday, 20 August, 2010, 13:14
> Dear Iasonas,
> 
> 
> > -----Original Message-----
> > From: r-help-bounces at r-project.org
> [mailto:r-help-bounces at r-project.org]
> On
> > Behalf Of Iasonas Lamprianou
> > Sent: August-20-10 5:55 AM
> > To: r-help at r-project.org
> > Subject: [R] Deviance Residuals
> > 
> > Dear all,
> > 
> > I am running a logistic regression and this is the
> output:
> > 
> > glm(formula = educationUniv ~ brncntr, family =
> binomial)
> > 
> > Deviance Residuals:
> >     Min   
>    1Q   Median   
>    3Q      Max  #
> αυτά είναι τα υπόλοιπα
> > -0.8825  -0.7684 
> -0.7684   1.5044   1.6516
> > 
> > Coefficients:
> >         
>    Estimate Std. Error z value Pr(>|z|)
> > (Intercept) -1.06869    0.01155
> -92.487   <2e-16 ***
> > brncntrNo    0.32654   
> 0.03742   8.726   <2e-16
> ***
> > ---
> > Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05
> '.' 0.1 ' ' 1
> > 
> > (Dispersion parameter for binomial family taken to be
> 1)
> > 
> > Null deviance: 49363  on 42969  degrees of
> freedom
> > Residual deviance: 49289  on 42968  degrees
> of freedom
> > AIC: 49293
> > 
> > 
> > I thought that the residuals should all be restricted
> in the range 0 to 1
> > (since I am predicting a binary outcome). I read many
> posts on this list and
> > I realized that there are four(!?) different types of
> residuals. I need a
> > simple account of these four types of residuals, if
> anyone can help it will
> > be great.
> > 
> > residuals(glm1, "response")
> 
> Residuals on the scale of the response, y - E(y); in a
> binary logistic regression, y is 0 or 1 and E(y) is the
> fitted probability of a 1. As it turns out, response
> residuals aren't terribly useful for a logit model.
> 
> > residuals(glm1, "pearson")
> 
> Components of the Pearson goodness-of-fit statistic.
> 
> > residuals(glm1, "deviance")
> 
> Components of the residual deviance for the model.
> 
> > residuals(glm1, "working") - especially this one
> confuses me a lot!
> 
> Residuals from the final weighted-least-squares regression
> of the IWLS procedure used to fit the model; useful, for
> example, for detecting nonlinearity.
> 
> > 
> > What is the "working" option and how is this
> different?
> 
> See above.
> 
> I hope this helps,
>  John
> 
> --------------------------------
> John Fox
> Senator William McMaster 
>   Professor of Social Statistics
> Department of Sociology
> McMaster University
> Hamilton, Ontario, Canada
> web: socserv.mcmaster.ca/jfox
> 
> > 
> > Thank you
> > Jason
> > 
> > Dr. Iasonas Lamprianou
> > 
> > 
> > Assistant Professor (Educational Research and
> Evaluation)
> > Department of Education Sciences
> > European University-Cyprus
> > P.O. Box 22006
> > 1516 Nicosia
> > Cyprus
> > Tel.: +357-22-713178
> > Fax: +357-22-590539
> > 
> > 
> > Honorary Research Fellow
> > Department of Education
> > The University of Manchester
> > Oxford Road, Manchester M13 9PL, UK
> > Tel. 0044  161 275 3485
> > iasonas.lamprianou at manchester.ac.uk
> > 
> > 
> > 
> > 
> > ______________________________________________
> > R-help at r-project.org
> mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained,
> reproducible code.
> 
> 
> 






More information about the R-help mailing list