I just developed a logistic regression model predicting customer churn (i.e how likely is a customer to leave us in the future?)

To understand the impact of my independent variables I calculated Odds ratio using the following function.


Where trainingmodel is the name of my model.

And I get the following results:



Both AIRTIME and Site Report are a feature of product we offer. In my dataset, AIRTIME is a continuous variable whereas Site.Report.By.Vehicle1 is a categorical variable with just two levels, ie. someone using the Site Report or not?

Can someone please help me to understand how to interpret the above number for AIRTIME and Site Report?

  • $\begingroup$ @Gregor don't think we can close with dupes from other sites. $\endgroup$ – David Arenburg Jun 11 '15 at 18:35
  • $\begingroup$ Maybe if we flag it a moderator can? Otherwise I think the most appropriate thing is to to migrate to Cross Validated and allow them to close as dup. $\endgroup$ – MrFlick Jun 11 '15 at 18:36
  • $\begingroup$ It looks like we can't close as other site dupe, but having the link in the comments will hopefully save some time at stats after migration. $\endgroup$ – Gregor Thomas Jun 11 '15 at 18:39
  • $\begingroup$ possible duplicate: How to interpret regression coefficients in logistic regression? $\endgroup$ – Gregor Thomas Jun 11 '15 at 21:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy