“…Unfortunately, the published literature addressing these two differing perspectives has been limited principally to open-label studies (33,34). The present investigation contained several important methodologic improvements: 1) the trial was blind, 2) it contained both an active-drug and a placebo comparison group, 3) the dosing permitted flexibility within a range in order to optimize each individual's dose, 4) the protocol excluded or controlled for other concomitant drug use, 5) it did not permit concurrent nonpharmacologic treatments, and 6) the data analysis strategy included a novel statistical method to differentiate direct versus indirect (or secondary) change in negative symptoms across the randomly assigned treatment groups.…”