UCLA Academic Technology Services HomeServicesClassesContactJobs
Stat Computing > SPSS > FAQ
 Search

SPSS FAQ
How do I test a group of variables in SPSS regression? 

Suppose that you want to run a regression model and to test the statistical significance of a group of variables.  For example, let's say that you want to predict students' writing score from their reading, math and science scores.  The data set with these variables in it can be downloaded by following this link:

hsb2.sav

The SPSS code for this would be:

regression
 /dependent = write
 /method = enter read math science.
Variables Entered/Removed(b)
Model Variables Entered Variables Removed Method
1 science score, reading score, math score(a) . Enter
a All requested variables entered.
b Dependent Variable: writing score
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .684(a) .467 .459 6.97111
a Predictors: (Constant), science score, reading score, math score
ANOVA(b)
Model Sum of Squares df Mean Square F Sig.
1 Regression 8353.990 3 2784.663 57.302 .000(a)
Residual 9524.885 196 48.596

Total 17878.875 199


a Predictors: (Constant), science score, reading score, math score
b Dependent Variable: writing score


Coefficients(a)

Unstandardized Coefficients Standardized Coefficients t Sig.
Model B Std. Error Beta
1 (Constant) 13.192 3.069
4.299 .000
reading score .236 .069 .255 3.410 .001
math score .319 .076 .316 4.222 .000
science score .202 .069 .211 2.918 .004
a Dependent Variable: writing score

Now let's suppose that you wanted to test the combined effect of math and science on writing.  The SPSS code for doing that is below.  Note that the variables listed in the /method = test() subcommand are not listed on the /method = enter subcommand.  In other words, the independent variables are listed only once.  Also note that, unlike other SPSS subcommands, you can have multiple /method = subcommands within the regression command. 

regression
 /dependent = write
 /method = enter read
 /method = test(math science).
Variables Entered/Removed(b)
Model Variables Entered Variables Removed Method
1 reading score(a) . Enter
2 science score, math score . Test
a All requested variables entered.
b Dependent Variable: writing score
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .597(a) .356 .353 7.62487
2 .684(b) .467 .459 6.97111
a Predictors: (Constant), reading score
b Predictors: (Constant), reading score, science score, math score
ANOVA(d)
Model Sum of Squares df Mean Square F Sig. R Square Change
1 Regression 6367.421 1 6367.421 109.521 .000(a)
Residual 11511.454 198 58.139


Total 17878.875 199



2 Subset Tests math score, science score 1986.569 2 993.284 20.439 .000(b) .111
Regression 8353.990 3 2784.663 57.302 .000(c)
Residual 9524.885 196 48.596


Total 17878.875 199



a Predictors: (Constant), reading score
b Tested against the full model.
c Predictors in the Full Model: (Constant), reading score, science score, math score.
d Dependent Variable: writing score


Coefficients(a)

Unstandardized Coefficients Standardized Coefficients t Sig.
Model B Std. Error Beta
1 (Constant) 23.959 2.806
8.539 .000
reading score .552 .053 .597 10.465 .000
2 (Constant) 13.192 3.069
4.299 .000
reading score .236 .069 .255 3.410 .001
math score .319 .076 .316 4.222 .000
science score .202 .069 .211 2.918 .004
a Dependent Variable: writing score




Excluded Variables(b)

Beta In t Sig. Partial Correlation Collinearity Statistics
Model Tolerance
1 math score .396(a) 5.583 .000 .370 .561
science score .322(a) 4.609 .000 .312 .603
a Predictors in the Model: (Constant), reading score
b Dependent Variable: writing score

If you wanted to test all three variables together, the synatx would be:

regression
 /dependent = write
 /method = test(read math science).
Variables Entered/Removed(a)
Model Variables Entered Variables Removed Method
1 science score, reading score, math score . Test
a Dependent Variable: writing score
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .684(a) .467 .459 6.97111
a Predictors: (Constant), science score, reading score, math score
ANOVA(c)
Model Sum of Squares df Mean Square F Sig. R Square Change
1 Subset Tests reading score, math score, science score 8353.990 3 2784.663 57.302 .000(a) .467
Regression 8353.990 3 2784.663 57.302 .000(b)
Residual 9524.885 196 48.596


Total 17878.875 199



a Tested against the full model.
b Predictors in the Full Model: (Constant), science score, reading score, math score.
c Dependent Variable: writing score


Coefficients(a)

Unstandardized Coefficients Standardized Coefficients t Sig.
Model B Std. Error Beta
1 (Constant) 13.192 3.069
4.299 .000
reading score .236 .069 .255 3.410 .001
math score .319 .076 .316 4.222 .000
science score .202 .069 .211 2.918 .004
a Dependent Variable: writing score

You will notice that the output from the first example with the three independent variables on the /method = enter subcommand and the output from this example with the three independent variables on the /method = test() subcommand are virtually identical.  The only difference between them is the line in the ANOVA table that gives the test of the subset, which in this case is all of the variables.  The point of this example is that you can put all of the independent variables in the regression on the /method = test() subcommand and not use a /method = enter subcommand if you like.

How to cite this page

Report an error on this page

UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services

The content of this web site should not be construed as an endorsement of any particular web site, book, or software product by the University of California