LOGISTIC REGRESSION Procedure
Unlike in SAS, the SPSS procedure LOGISTIC REGRESSION models the probability of Y=1 or Y's higher sorted value. Suppose the response variable Y is 0 or 1 binary (This is not a limitation for SPSS either. The values can be either numeric or character as long as they are dichotomous), and X1 and X2 are two regressors of interest. To run a logistic regression, use:
logistic regression var=y with x1 x2.
Using the data in Example 1, you can use:
logistic regression var=s with t.
You will have the SPSS output:
L O G I S T I C R E G R E S S I O N
Total number of cases: 387 (Unweighted)
Number of selected cases: 387
Number of unselected cases: 0
Number of selected cases: 387
Number rejected because of missing data: 0
Number of cases included in the analysis: 387
Dependent Variable Encoding:
Original Internal
Value Value
.00 0
1.00 1
Dependent Variable.. S
Beginning Block Number 0. Initial Log Likelihood Function
-2 Log Likelihood 106.98843
* Constant is included in the model.
Beginning Block Number 1. Method: Enter
Variable(s) Entered on Step Number
1.. T
Estimation terminated at iteration number 6 because
Log Likelihood decreased by less than .01 percent.
-2 Log Likelihood 95.375
Goodness of Fit 346.446
Chi-Square df Significance
Model Chi-Square 11.614 1 .0007
Improvement 11.614 1 .0007
Classification Table for S
Predicted
.00 1.00 Percent Correct
0 I 1
Observed +-------+-------+
.00 0 I 0 I 12 I .00%
+-------+-------+
1.00 1 I 0 I 375 I 100.00%
+-------+-------+
Overall 96.90%
---------------------- Variables in the Equation -----------------------
Variable B S.E. Wald df Sig R Exp(B)
T -.0807 .0224 13.0289 1 .0003 -.3211 .9225
Constant 5.4152 .7275 55.4000 1 .0000
The output shows that the estimated logit is
where p is the probability of having an ingot ready for rolling. This is the same result as with the use of the DESCENDING option in SAS PROC LOGISTIC.
PROBIT Procedure
You can also use the SPSS PROBIT procedure to fit a logistic regression. The PROBIT procedure supports the model with grouped data. To fit a logistic regression, use:
probit r of n with x1 x2 /model logit.
where R represents the response count and N represents the observation count.
Using the same data in Example 1, you can use the following syntax. Notice that N=1 must be generated for all the observations because the number of trials is 1 for this individual data set.
compute n = 1. execute. probit r of n with t /model logit /print none.
The resulting SPSS output will be:
* * * * * * * * * * * * P R O B I T A N A L Y S I S * * * * * * * * * * * *
Parameter estimates converged after 16 iterations.
Optimal solution found.
Parameter Estimates (LOGIT model: (LOG(p/(1-p))) = Intercept + BX):
Regression Coeff. Standard Error Coeff./S.E.
T -.08070 .02236 -3.60965
Intercept Standard Error Intercept/S.E.
5.41524 .72755 7.44314
Pearson Goodness-of-Fit Chi Square = 346.453 DF = 385 P = .921
Since Goodness-of-Fit Chi square is NOT significant, no heterogeneity
factor is used in the calculation of confidence limits.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Next: Probit Regression
Prev: Logistic Regression with SAS
Up: Contents
