Html Version of SPSS Code to run exemplar 5 weighting 
It is intended to show you the code and to allow links, not to use
as an SPSS syntax file. 
The SPSS syntax is in  is the file ex5wt.sps which you should
save to a file and/or read in to SPSS.
To see output from the commands go to the SPSS results file.


Links in this page
With continuous age groups
Getting the predictor and the weights


back to top 

GET FILE='C:\Documents and Settings\gillian raab\My Documents\aprojects\peas\web\exemp5\data\ex5wt.sav'.

*  Line above reads in the data. If you have opened ex5wt.sav  interactively you won't need to include it.
*---------------------------------------------------------------------------------------------------------------------------------------.
*logistic regression with grouped data in SPSS can only be done using PROBIT.
*Only the final model is shown here, although otehrs could have been tried.

*first compute interaction terms.
*---------------------------------------------------------------------------------------------------------------------------------------------.
compute agegrp2=agegrp*agegrp.
compute agesinc=agegrp*sinc.
compute sex=0.
if gender="F" sex=1.
compute agesex=agegrp*sex.
compute age2sex=agegrp2*sex.
execute.
*------------------------------------------------------------------------------------------------------------------------.
*  Now the logistic regression.
* needed extra iterations to converge.
*  this model is not quite the same as SAS since the agegp quadratic interaction with sinc.
* has been missed out.
* ------------------------------------------------------------------------------------------------------------------------..

PROBIT
  AyrA_num  OF npop  WITH sinc sacc  agegrp sex agegrp2  agesex age2sex agesinc
  /LOG NONE
  /MODEL LOGIT
  /CRITERIA ITERATE(50) STEPLIMIT(.1) .
*-------------------------------------------------------------------------------------
back to top 

*The PROBIT command doesn't give predicted values in a file.
*although it has an option to print them out.
*so it is necessary to compute them from the output.
*--------------------------------------------------------------------------------------------------------- .
   compute lp= -4.47972-0.02489*sinc+sacc*0.04634+agegrp*(-0.00248)   +
    sex*.47620 + agegrp2  *(.00721)  +agesex*(-.00720) +age2sex*(-.00508)+agesinc*(.00152).  
compute phat=1/(1+exp(-lp)). 
compute gweight=1/phat.
execute. 
compute weighta=gweight/75.60964.