Confidence Intervals for Predicted Probabilities

[Duwe, Grant (DOC)]

Is anyone aware of any available syntax, macro, or other procedure for SPSS that computes confidence intervals for predicted probabilities saved from a logistic regression model? I’m aware that other packages such as Stata have this capability, but I’d prefer to find a solution in SPSS.

 

Thanks,

GD

---

This email is intended to be read only by the intended recipient. This email may be legally privileged or protected from disclosure by law. If you are not the intended recipient, any dissemination of this email or any attachments is strictly prohibited, and you should refrain from reading this email or examining any attachments. If you received this email in error, please notify the sender immediately and delete this email and any attachments. Thank you.

[Bruce Weaver]

Duwe, Grant (DOC) wrote

Is anyone aware of any available syntax, macro, or other procedure for SPSS that computes confidence intervals for predicted probabilities saved from a logistic regression model? I'm aware that other packages such as Stata have this capability, but I'd prefer to find a solution in SPSS.

Thanks,
GD

Have you tried running your logistic regression model with GENLIN (with binomial error distribution and logit link function)?  GENLIN has more options for saved variables.  From a quick look at the Help, I think it will give save the predicted log-odds with associated confidence interval.  You should be able to convert those to predicted probabilities (and CI) if that's what you want.

HTH.