I am unable to find the "secret" to getting output with R-squared changes

when running a stepwise regression. I've tried the Help, syntax commands,

etc., but haven't had any success. Otherwise, I seem able to do whatever I

want with SPSS, but I'd like to see my R-squared changes. Suggestions,

please? Thanks, Marjorie

Just use the CHA keyword on the statistics subcommand in Regression.

David Futrell

Indianapolis

> I am unable to find the "secret" to getting output with R-squared changes

> when running a stepwise regression. I've tried the Help, syntax commands,

> etc., but haven't had any success. Otherwise, I seem able to do whatever I

> want with SPSS, but I'd like to see my R-squared changes. Suggestions,

> please? Thanks, Marjorie

Within your Regression command, add the subcommand "/statistics=cha" .

This will give you r-squared change stats.

Jennifer R. Popovic

Statistician

Surveys and Research Staff

Veterans Benefits Administration

202.273.6463 (phone)

202.275.5947 (FAX)

>----------

>Sent: Tuesday, September 08, 1998 4:06 PM

>Subject: R-squared changes

>Sent: Tuesday, September 08, 1998 4:06 PM

>Subject: R-squared changes

>I am unable to find the "secret" to getting output with R-squared changes

>when running a stepwise regression. I've tried the Help, syntax commands,

>etc., but haven't had any success. Otherwise, I seem able to do whatever I

>want with SPSS, but I'd like to see my R-squared changes. Suggestions,

>please? Thanks, Marjorie

> I am unable to find the "secret" to getting output with R-squared changes

> when running a stepwise regression. I've tried the Help, syntax commands,

> etc., but haven't had any success. Otherwise, I seem able to do whatever I

> want with SPSS, but I'd like to see my R-squared changes. Suggestions,

> please? Thanks, Marjorie

7.5, there is a "statistics" button at the bottom of the regression dialog

box. If you click on that button, you will see the "r squared change"

option at the upper right... click in the box and back out...

__________________________________________________________________________

William B. Ware, Professor and Chair Psychological Studies

CB# 3500 in Education

University of North Carolina PHONE (919)-962-7848

Chapel Hill, NC 27599-3500 FAX: (919)-962-1533

__________________________________________________________________________

1. Comparing OLS R-squared with McFadden R-squared (ORDINAL REGRESSION)

I did a study using Ordinal Regression in SPSS, but virtually all

previous similar studies were done on the exact same type of data

(discrete ordinal categories) using OLS. The OLS R2 values in these

other studies are much higher than the McFadden pseudo-R2 values I get

out. I wondered if the pseudo-R2 value was truly comparable to the OLS

R2 values, and if not, why not? Is a McFadden pseudo-R2 of 0.2 as good

or better than as an OLS R2 of 0.4?

I would be extremely appreciative for any pointers to simple (few

formulae, small words, something economists can understand) discussions

of the quality (superiority or inferiority) between OLS R2 and McFadden

pseudo R2, or lack of such differences, as measures of overall model fit.

Cheers -julii

Dr. J. Brainard UEA Norwich, UK

2. @home/E-Mail/Linksys Router Problem

3. R-squared

4. Can't Create Tables, Views , SP from the Server manager

6. Beginner checking for memory leaks

7. (pseudo) r-squared for PROC MIXED

8. Syslog to named pipe/fifo help needed

9. pseudo r-squared in logistic regression

10. r-squared in proc logistic

12. no-intercept model and R-squared

13. R-squared in logistic regression

4 post • Page:**1** of **1**