The Authors Respond

We have mixed feelings about this feedback. On the one hand, we acknowledge that the number of events per variable for some of the logistic regression analyses, particularly the one to predict a final diagnosis of major depression on the basis of demographic and injury variables, was suboptimal. This was because of the relatively low number of participants (N=24, 17%) with such a diagnosis. We could have addressed this by limiting the number of predictors in the model but decided against this because (1) they had been selected a priori for theoretical reasons, and (2) the participant-to-variable ratio in the model was acceptable.
Source: Archives of Physical Medicine and Rehabilitation - Category: Rehabilitation Authors: Tags: Departments Source Type: research