Testing for Regression Model Assumptions

[Michael Maley]

How do I check for linear regression model assumptions (linearity,
homoscedasticity, normality) in SPSS?

Thanks,
Michael

[Maguin, Eugene]

Michael,

In a few words, look at the residuals via the residuals, casewise,
scatterplot and partialplot commands. For a more serious analysis you can
use the save command to save various residuals into your working datafile
and then use the examine procedure. Write back if you want more info.

Gene Maguin

[Hector Maletta]

Gene Maguin's advice is right, only those are SUBcommands of the REGRESSION
command. You may want to look at Marija Norusis' SPSS Guide to data analysis
for specific advice on analyzing regression residuals in SPSS.

Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de Gene
Maguin
Enviado el: Wednesday, June 21, 2006 11:22 AM
Para: [hidden email]
Asunto: Re: Testing for Regression Model Assumptions

Michael,

In a few words, look at the residuals via the residuals, casewise,
scatterplot and partialplot commands. For a more serious analysis you can
use the save command to save various residuals into your working datafile
and then use the examine procedure. Write back if you want more info.

Gene Maguin

[paulandpen]

Try the following link below, this might help, if you can get over the accent (hehehehe)

Regards Paul

http://distdell4.ad.stat.tamu.edu/spss_1/RegDiagnostics.html




> Hector Maletta <[hidden email]> wrote:
>
Gene Maguin's advice is right, only those are SUBcommands of the REGRESSION
command. You may want to look at Marija Norusis' SPSS Guide to data analysis
for specific advice on analyzing regression residuals in SPSS.

Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de Gene
Maguin
Enviado el: Wednesday, June 21, 2006 11:22 AM
Para: [hidden email]
Asunto: Re: Testing for Regression Model Assumptions

Michael,

In a few words, look at the residuals via the residuals, casewise,
scatterplot and partialplot commands. For a more serious analysis you can
use the save command to save various residuals into your working datafile
and then use the examine procedure. Write back if you want more info.

Gene Maguin

[Kaukab Azim]

Hi Paul,
  Thanks for sending the link. Are there any more of such tutorials?
  Kaukab.

Paul Dickson <[hidden email]> wrote:
  Try the following link below, this might help, if you can get over the accent (hehehehe)

Regards Paul

http://distdell4.ad.stat.tamu.edu/spss_1/RegDiagnostics.html




> Hector Maletta wrote:
>
Gene Maguin's advice is right, only those are SUBcommands of the REGRESSION
command. You may want to look at Marija Norusis' SPSS Guide to data analysis
for specific advice on analyzing regression residuals in SPSS.

Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de Gene
Maguin
Enviado el: Wednesday, June 21, 2006 11:22 AM
Para: [hidden email]
Asunto: Re: Testing for Regression Model Assumptions

Michael,

In a few words, look at the residuals via the residuals, casewise,
scatterplot and partialplot commands. For a more serious analysis you can
use the save command to save various residuals into your working datafile
and then use the examine procedure. Write back if you want more info.

Gene Maguin




Dr. Kaukab Azim

Room #: 1.14 (The Genetics Lab)

Institute of Molecular Medicine

St. James's Hospital

Dublin 8

Ireland.





---------------------------------
Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates.