I am looking at a results chart from a binomial logistic regression model, and I see the following results. I am not sure what "intercept" means here. I thought "coefficient" is referring to the variable in question - so it makes no sense to have no coefficient in the first line , enter image description here Also, is "Downtown" location being used as a reference to check the results between the two locations?


"Ref" is shorthand for reference category. Whenever an independent variable consists of categories, logistic regression procedures (if using this particular sort of contrast -- an indicator contrast) will choose one category by default as the reference to which each other category is compared. Here, the odds of the outcome being "1" are estimated as being lower (0.57 as high) for North cases as for Downtown cases, all else being equal.

The intercept (also known as the regression constant) tells you nothing about the role of any independent variable, but it does serve a function in a predictive equation that can be derived from output like this. You can use the constant and the two other coefficients to assign to each observation a predicted probability of the outcome being "1", depending on that case's values for the independent variables.

  • $\begingroup$ I wish I could vote for your answer but the site doesn't allow me given I don't have enough reputation. Thank you, this really helped! $\endgroup$ – BitwiseStarter Feb 7 '18 at 2:54
  • $\begingroup$ what do you mean by "estimated as being lower (0.57 as high) for North cases as for Downtown cases". So is it that the odds of being in North is 0.57? I've been looking into odds ratio, and for some reason, ALL examples I see starts with 1.xx, 1.57, etc. What does it mean by it's 0.xx? $\endgroup$ – BitwiseStarter Feb 8 '18 at 2:51

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.