I want to run a probit or a logit model and I am curious about the choise regarding the data sample size that I have. In a previous answer in a question 'Probit vs Logit' I read that "Probit is better in the case of "random effects models" with moderate or large sample sizes (it is equal to logit for small sample sizes)". How can someone categorize the data sample as small, moderate or large? For example a dataset of 5000 observations can be identifed as small, moderate or large ?
Asked
Active
Viewed 29 times
1
-
"Small," "moderate," and "large" have no meaning outside of some comparative context. Is a weight of 6,000 kilos "small," for instance? It depends on whether you are talking about people, dinosaurs, or mountains. Could you provide a link to the "previous answer" so we might guess at the context? – whuber Feb 14 '22 at 20:09
-
Here is the link: https://stats.stackexchange.com/questions/20523/difference-between-logit-and-probit-models – Collin Focas Feb 15 '22 at 07:45
-
Thank you. You might be reading too much into a very specific comment (about probit vs. logit models) within an obscure answer. It's possible you will get a better response by asking the person who posted that answer what they had in mind. – whuber Feb 15 '22 at 14:38
-
Thank you, but I can't ask the person as I am a new member. The system does not allow me to write comments. – Collin Focas Feb 15 '22 at 14:56
-
@CollinFocas Who asked, and in response to which post? I can post a comment. You'll be able to post comments before too long, too. (I remember the early days when I couldn't comment. It didn't last long.) – Dave Feb 15 '22 at 15:42
-
It was this post: https://stats.stackexchange.com/questions/20523/difference-between-logit-and-probit-models , you can scroll down to find the answer in which I am curious about – Collin Focas Feb 15 '22 at 15:44
-
Who posted the comment to which you'd like to respond, and to whose post was it a comment? – Dave Feb 15 '22 at 15:46
-
This person posted the comment https://stats.stackexchange.com/users/81392/tripartio, to this post: https://stats.stackexchange.com/questions/20523/difference-between-logit-and-probit-models – Collin Focas Feb 15 '22 at 15:47