What is the most appropriate (a priori) methodology to address sample size in an observational study analyzing risk factors and injury outcomes?

Acknowledging the multiple methodologies available (and opinions on the subject), what is the most robust method while also being the most practical?

09 March 2013 8,366 2 View