I applaud the Walach et al for their attempts at applying quantitative methods to qualitative research1. Having myself conducted 70 homeopathic drug provings following GCP guidelines, one with 110 subjects and an equal placebo group, I view this research as important in trying to evaluate whether or not it is possible, in a proving, to distinguish reliable symptoms specific to a medicine from random, non-specific symptoms. Many well-proven and commonly prescribed homeopathic remedies have ‘keynotes’ that did not appear in a homeopathic drug proving. However, I have some questions. Since we know that placebo symptoms in conventional medical research vary significantly across clinical trials for the same disease and also among different organ systems I am unclear how the authors concluded that what was experienced during the proving period was different from background noise...