We agree with Drs Batterham and Hopkins that small studies are not necessarily inherently flawed. However, probabilistically speaking, even in the absence of biases, small-sized trials are more prone to provide overestimates (or underestimates) compared with larger trials. Thus, evidence from scattered small studies is easier to distort than evidence from large trials because analyses with the most impressive results are more likely to be published compared with studies showing underestimated treatment effects.
展开▼