«Such
biases, conscious or unconscious, can make candidate medical treatments look better than they actually are, the authors of the analysis warn, and lead to eye - catching results that can't be
replicated in larger or more rigorous animal studies — or in
human trials.»