should we include or exclude a variable in a logit regr. model which will only obtain values if a certain event takes place otherwise will show N/A? this variable tells whether or not a product will be bought based on calls made by the company. the variable tells: number of days that passed by after the client was last contacted from a previous campaign. the values (-1) are for users who were not previously contacted enter image description here

  • 1
    $\begingroup$ It could be possible that the not being contacted has an effect on the response. As a result, I wouldn't suggest dropping the variable. You could try transforming it into categories (e.i. [contacted (Yes/No)] or [contacted (No, 0-100 days, 100-200 days, 200-300 days, 300+ days)] - these are two possible options, out of many you, could explore. $\endgroup$
    – nwaldo
    Apr 19, 2020 at 19:18
  • $\begingroup$ adding to what @nwaldo said, you may also try taking zero as a value for "No", in which case the effect on the resulting regression is likewise zero $\endgroup$ Apr 19, 2020 at 21:53
  • $\begingroup$ @AndreasSteimer thank you for the help. also wanted to ask if the changing to numerical forms is necessary before fitting the logit reg model?stats.stackexchange.com/questions/461535/… $\endgroup$
    – shrippi
    Apr 19, 2020 at 23:34


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.