On 7 Apr 2001 11:16:49 -0700 [EMAIL PROTECTED] (Dianne Worth)
wrote:
> After several years of frustration with SAS, I am migrating to SPSS.
> I am currently working on a project in both packages, to ensure
> accuracy of results as I teach myself SPSS. I would like to
> obtain 1) the squared semi-partial correlation based on the sequence
> that predictors are entered into the model statement (SCORR1 in SAS)
> and 2) SCORR2, which is supposed to show the unique proportion of
> variance that the predictor explains in Y.
>
> Can anyone tell me how to do this in SPSS? I tried
> ANALYZE/CORRELATIONS/PARTIAL but the answers were different than
> the SAS output.
It's easy to compute partial correlations and semipartial correlations
in either SAS or SPSS. Here is a sample demonstration program
in SAS for computing semipartials that I wrote for someone else.
**********
hdata:
4 2 4
3 4 1
6 6 5
1 1 2
5 3 3
program:
OPTIONS PS=60 LS=80;
COMMENT PROGRAM TO DEMONSTRATE THE COMPUTATION OF SEMIPARTIAL
COMMENT CORRELATION COEFFICIENTS USING SAS PROC CANCORR;
COMMENT CAROL NICKERSON / 30 DECEMBER 1999;
DATA TEMP1;
INFILE 'hdata' PAD;
INPUT Y X1 X2;
PROC PRINT DATA = TEMP1;
PROC REG NOPRINT DATA = TEMP1;
MODEL X1 = X2;
OUTPUT OUT = TEMP2 R = RESID1;
PROC REG NOPRINT DATA = TEMP2;
MODEL X2 = X1;
OUTPUT OUT = TEMP3 R = RESID2;
PROC CORR NOSIMPLE NOPROB DATA = TEMP3;
WITH Y;
VAR RESID1 RESID2;
PROC REG DATA = TEMP3;
MODEL Y = X1 X2 / SCORR2;
PROC CANCORR NOPRINT SPCORR SQSPCORR DATA = TEMP3 OUTSTAT = TEMP4;
VAR Y;
WITH X1 X2;
DATA TEMP5;
SET TEMP4;
IF (_TYPE_ EQ 'SPCORR') OR (_TYPE_ EQ 'SQSPCORR');
IF _NAME_ EQ 'Y';
PROC PRINT DATA = TEMP5;
VAR _TYPE_ X1 X2;
**********
And here is a demonstration program in SPSS. SPSS calls semipartial
correlations by the alternative name "part correlations".
**********
DATA LIST FREE /
Y X Z
BEGIN DATA
4 2 4
3 4 1
6 6 5
1 1 2
5 3 3
END DATA
LIST CASES
REGRESSION / VARS Y X
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER
/ WIDTH = 80
REGRESSION / VARS Y Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER
/ WIDTH = 80
REGRESSION / VARS X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=X
/ METHOD = ENTER
/ WIDTH = 80
REGRESSION / VARS Y X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER
/ WIDTH = 80
REGRESSION / VARS Y X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER X Z
/ WIDTH = 80
REGRESSION / VARS Y X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER Z X
/ WIDTH = 80
REGRESSION / VARS Y X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER Z
/ METHOD = ENTER X
/ WIDTH = 80
REGRESSION / VARS Y X Z
/ DESCRIPTIVES=ALL
/ STATISTICS=ALL
/ DEPENDENT=Y
/ METHOD = ENTER X
/ METHOD = ENTER Z
/ WIDTH = 80
**********
The book by Cody and Smith is a very useful SAS reference.
All the best,
Carol Nickerson
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
