CSPLAN in SPSS vs STATA

[Mirka]

Dear All,
I have been trying to replicate STATA  set-up for complex sample
analysis - without any sucess:
svyset unique_HHLD_ID [pw=NADULTS], strata (unityear)
postweight(Tot_popn_STATSCAN) poststrata(unityear)
- without any sucess.
Here are my tries with SPSS:
* Analysis Preparation Wizard.
CSPLAN ANALYSIS
  /PLAN FILE='G:\HHLDandPopn_weights_2.csaplan'
  /PLANVARS ANALYSISWEIGHT=NADULTS
  /PRINT PLAN
  /DESIGN STRATA=unityear
  /ESTIMATOR TYPE=EQUAL_WOR
  /POPSIZE VARIABLE=Tot_popn_STATSCAN.

* Analysis Preparation Wizard.
CSPLAN ANALYSIS
  /PLAN FILE='G:\HHLDandPopn_weights_2.csaplan'
  /PLANVARS ANALYSISWEIGHT=NADULTS
  /PRINT PLAN
  /DESIGN STRATA=unityear
  /ESTIMATOR TYPE=EQUAL_WOR
  /POPSIZE VARIABLE=Tot_popn_STATSCAN.

Both plans  are producing the same output, the the mean estimates are
the same as in STATA, but the standard errors are smaller in SPSS.

Any help would be appreciate,
thank you,
mirka

--

Mirka Ondrack                     Tel:[phone][ex:20306]

Institute for Social Research     Fax:[phone]

York University, TEL 5063         email: [hidden email]
<mailto:[hidden email]>

Statistical Consulting Services:  http://www.isr.yorku.ca/scs/index.html

Online Appointment Scheduler:
http://www.appointmentquest.com/provider/2000199121

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD