Logistic Regression on Aggregate Data Help!

[DemPyro]

Hello,

I'm trying to perform an adjusted logistic regression. My data set is as follows:

Sex DR Anx Count
Female .00 .00 509615.00
Female .00 1.00 22717.00
Female 1.00 .00 2064.00
Female 1.00 1.00 621.00
Male .00 .00 424300.00
Male .00 1.00 11171.00
Male 1.00 .00 1286.00
Male 1.00 1.00 235.00

I'd like to be able to test the dependence of the Anx outcome due to Sex and DR. The count is the number of cases appeared for each result. How would I go about running such an analysis? Is there a different way for me to set up the data to be able to run this analysis?


Thanks

[Bruce Weaver]

Post by DemPyro
Hello,
Sex DR Anx Count
Female .00 .00 509615.00
Female .00 1.00 22717.00
Female 1.00 .00 2064.00
Female 1.00 1.00 621.00
Male .00 .00 424300.00
Male .00 1.00 11171.00
Male 1.00 .00 1286.00
Male 1.00 1.00 235.00
I'd like to be able to test the dependence of the Anx outcome due to Sex and DR. The count is the number of cases appeared for each result. How would I go about running such an analysis? Is there a different way for me to set up the data to be able to run this analysis?
Thanks

Look up the WEIGHT command.

https://www.ibm.com/support/knowledgecenter/en/SSLVMB_25.0.0/statistics_reference_project_ddita/spss/base/syn_weight.html

Here is an example using your data. Notice that I changed the variable showing the cell counts from Count to N (because Count is the name of a command in SPSS).

NEW FILE.
DATASET CLOSE ALL.
DATA LIST LIST / Sex (A6) DR Anx (2F1) N (F8.0).
BEGIN DATA
Female .00 .00 509615.00
Female .00 1.00 22717.00
Female 1.00 .00 2064.00
Female 1.00 1.00 621.00
Male .00 .00 424300.00
Male .00 1.00 11171.00
Male 1.00 .00 1286.00
Male 1.00 1.00 235.00
END DATA.

WEIGHT by N.

LOGISTIC REGRESSION VARIABLES Anx
/METHOD=ENTER DR Sex
/CONTRAST (Sex)=Indicator
/CONTRAST (DR)=Indicator
/PRINT=CI(95)
/CRITERIA=PIN(0.05) POUT(0.10) ITERATE(20) CUT(0.5).