*(Q) Tom Ryan's recent book chapter on logistic regression (1997; Modern Regression 
Methods) lists the Wald test as a 'poorly' performing test of individual predictors. 
Ryan goes on to describe 3 other tests: 
	the likelihood ratio test, 
	the scores test, and 
	the exact conditional scores test. 
Apparently, SPSS 11.5 offers only the Wald test. Does any one have any input on this?

*(A) Posted to SPSSX-L list on 2002/12/23 by Marta Garcia-Granero.

To get the individual p values (either likelihood ratio test or exact conditional 
scores test), try this trick:

LOGISTIC REGRESSION VAR="your dependent variable"
  /METHOD=BSTEP(LR) "list of IV"
or
  /METHOD=BSTEP(COND) "list of IV"
.
(Rest of options come here)
.
  /CRITERIA PIN(1) POUT(1).

As you can see, you select backwards elimination of the IV, but forcing a PIN and 
POUT value of 1 (you don't eliminate anything from the model). This is the way of 
forcing SPSS give the individual p-values for the likelihood ratio test (BSTEP(LR)),
or conditional test (BSTEP(COND)).