Why it is recommended not to use continuous form variable as independent variable in logistic regression?

I know of course only categorial form data can be used as dependent variable, and when it comes to independent variable all kinds of forms can be used.
But I saw somewhere that it is not recommended to use continuous form as independent variable and rather transform it into categorial or binomial form.
Can anyone tell me the reason of this?

T'aint so. Continuous variables are used in logistic regression all the time.

Are you referring to binning and other tools that take continuous data and make them discrete?

If that's the case, take a look at Discretize Predictors as a Last Resort or just about anything that Frank Harrell has published. There's a lot of literature about this (see the references in the link above). binning can simultaneously reduce power to find real patterns as well as increase the risk of false detection of non-existent patterns.

It does happen all of the time though :crying_cat_face:

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.