I have a model in which I suspect two variables to may be related. However, one is a binary/dichotomous variable, while the other is a continous variable. What approach should be used to check if my statement has any merit or not? The Point-Biserial Correlation test, or...?
$\begingroup$
$\endgroup$
Add a comment
|
$\begingroup$
$\endgroup$
Yes, point-biserial correlation is an option. You can also run a t-test and compare means between the two levels of your binary variable. These two approaches actually are algebraically identical (Ref).
