Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: st: RE: Interpretation of quadratic terms
From
"Lachenbruch, Peter" <[email protected]>
To
"'[email protected]'" <[email protected]>
Subject
RE: st: RE: Interpretation of quadratic terms
Date
Mon, 8 Mar 2010 15:57:01 -0800
A couple of posts noted a) using residuals after logit or logistic gives you Pearson residuals (obs-expected)/sqrt(var), b) after glm you get observed -expected (and get the two straight lines).
Check out the postestimation stuff for logit and glm and you'll see what I mean. Also, try this on the auto data - e.g. predicting foreign.
Tony
Peter A. Lachenbruch
Department of Public Health
Oregon State University
Corvallis, OR 97330
Phone: 541-737-3832
FAX: 541-737-4001
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Rosie Chen
Sent: Monday, March 08, 2010 2:54 PM
To: [email protected]
Subject: Re: st: RE: Interpretation of quadratic terms
Thank you all for the very helpful input and advice.
It seems that -glm- has a general approach. It would be great if I can find how to check residual after the -logit-syntax. Will dig further and see how it goes. Otherwise, will use the glm approach. Thank you again!
Rosie
----- Original Message ----
From: Richard Williams <[email protected]>
To: "[email protected]" <[email protected]>; "[email protected]" <[email protected]>
Sent: Mon, March 8, 2010 5:36:12 PM
Subject: RE: st: RE: Interpretation of quadratic terms
At 05:22 PM 3/8/2010, Martin Weiss wrote:
><>
>
>" I think that is why you don't even get the usual resid = y
>- yhat as an option with the predict command after logit:"
>
>
>
>How about the ones available after -glm-?
-glm- has more of a one size fits all approach. It doesn't have a
lot of the customizations you see when using the more specific
commands. There was a thread a long time ago about how after -logit-
some residuals are computed "adjusted for number sharing covariate
pattern" -- if that isn't the way you want to compute them you can
use -glm- instead. But logit restricts you the way it does because
the options listed are the ones deemed more statistically sound.
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME: (574)289-5227
EMAIL: [email protected]
WWW: http://www.nd.edu/~rwilliam
*
* 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/
*
* 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/