Logistic regressions with small n's in Yes categories

[Judith Saebel]

Dear All,

 

When running a logistic regression, what is the smallest acceptable
number in the yes category?  We need to run a few analyses, but the
problem is that in some of them, the ratio Yes/No is something like
8/200.

 

Thanks in advance for any ideas.

 

Sincerely,

 

 

Judith Saebel