I am running a binary logistic regression for my data using mini-tab. I feel as though I am doing something wrong because the p-values given for each of binary categories for an independent variable are the exact same number. But, this doesn't make sense as the numerical inputs are not the same.

For example: Male to Female the p value is exactly the same But, when looking at the actual data counts input it is as follows: Male: 210 likely, 41 not likely Female: 239 likely, 20 not likely

While both male and female are the same independent variable (gender) I feel like I must be doing something wrong because how could those p-values be the same in relation to significance of predictability of an event to likely occur or not when the values are not identical.

enter image description here

enter image description here

  • $\begingroup$ Are you including “male” and “female” in your model as separate variables? It would be typical to include one variable indicating gender, coding males as $0$ and females as $1$ or vice versa. If you include both, then your variables contain identical information. $\endgroup$
    – Dave
    Nov 25, 2021 at 16:08
  • $\begingroup$ I have M/F as one variable. I have one column labeled as gender and had coded it initially as M and F within that column. I've added a picture to my initial post to show how the data looks. (note: this edited data has the inputs as 0 and 1 not M and F) $\endgroup$
    – Sarah
    Nov 25, 2021 at 16:51


Your Answer

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