277 followers
An interesting new method - it tests the continuous outcome for associations first, before testing multiple cut-offs to generate RR in post-hoc analyses. Avoids the dilemma in choosing cut-off and dichotomizing a continuous outcome from the beginning! http