I have a categorical data on gender with a sample size of over 20,000 datasets. In what cases can one use a probit regression over a logit regression?
IMO NOT much difference. I use logistic regression unless I have a particular reason to do something else
See attached paper as an example and the second attachment for the exact difference. Best wishes David Booth
This is will help you
https://tutorials.methodsconsultants.com/posts/what-is-the-difference-between-logit-and-probit-models/
Have a look at my quite lengthy answer to this
https://www.researchgate.net/post/How_to_decide_on_which_model_Logit_or_Probit_to_use_when_modeling_the_business_cycle_phase
IMO it depends on your audience and what you are trying to do. In the public health area, logit is usually used because of the odds ratio. The probit is generally used in econometrics.
I'll add one more bit into the mix. this is one way a biostatistician may view the choice as described in the screenshot. Best wishes David Booth
Logistic regression model is most preferred because of the features I earlier explained to you.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression