1

I am Looking for the Reason That we do not Need Normal distribution on Independent and dependent variables in Logistic Regression, can anybody explain it ?

kjetil b halvorsen
  • 77,844
nasim_bb
  • 21
  • 2
    Welcome to CV. Since you’re new here, you may want to take our tour, which has information for new users. Since this looks like homework (apologies if it's not), please add the [self-study] tag and read its wiki. Then tell us what you understand thus far, what you've tried & where you're stuck. We'll provide hints to help you get unstuck. Please make these changes as just posting your homework & hoping someone will do it for you is grounds for closing. If this is self-study rather than homework, let us know, and... it's still a good idea to show us what you've tried. – jbowman Dec 20 '22 at 18:00
  • 1
    Your phrasing suggests to me that believe these to be needed in other forms of regression, such as linear regression. If you do believe this, it would be helpful to say what you believe about that, as there are common misconceptions surrounding these, and I expect it to be easier to believe that normality is not necessary for logistic regression once you believe it it not necessary for linear regression. – Dave Dec 20 '22 at 18:07
  • 2
    Because the dependent variable in logistic regression can, by definition, take on only two possible values, a Normal distribution would be a spectacularly poor model for its conditional distribution. – whuber Dec 20 '22 at 18:50
  • @jbowman Thank You. No It is not for my homework, i am just want to work on logestic regression and now i am reading about it. i read about LR assumption and every where people said that LR don't need any normality (not in independent variable, not in dependent and not in error term) but no where they explain about whay LR dont need normality and because i am new to this things i can't undrestand why. – nasim_bb Dec 20 '22 at 19:20
  • @Dave no, i read that linear regression also don't need normality in independent and dependent variable ....but because i want to work on LR, I ask about that...just it – nasim_bb Dec 20 '22 at 19:22
  • @whuber Thank you...but how about independent variables? – nasim_bb Dec 20 '22 at 19:23
  • 2
    Things to ponder: 1) Why would they require normality? After all, we can ask why don't we need Cauchy independent variables, Laplace independent variables, or... just as easily as Normal independent variables. 2) Think of sex (0/1) as an independent variable - would it be invalid to have sex as an independent variable because it's not Normally distributed? – jbowman Dec 20 '22 at 19:29
  • Regression concerns the distribution of the response conditional on the explanatory variables. That neither requires nor makes assumptions about the distribution of the explanatory variables. It can sometimes help to re-express the explanatory variables, though, because that can make the regression linear. See https://stats.stackexchange.com/a/35717/919 for a worked example. The same principles apply in logistic regression. A simple way to check the data for a linear relationship is given at https://stats.stackexchange.com/a/14501/919. – whuber Dec 20 '22 at 22:29

0 Answers0