The Perils of 'Bite Size' Science - NYTimes.com - 1 views
-
Ryan Catalani on 29 Jan 12I've noticed this trend, too, in a lot of the articles I've posted recently. Something to be aware of: "...we worry that shorter, single-study articles can be poor models of science. ... we are troubled by the link between small study size and publication bias. Theoretically, if several small studies on a topic, each with its own small data set, are sent to publishers, the overall published results should be equivalent to the results of a single large study on that topic using a complete data set. But according to several "meta-studies" that have been conducted, this is often not the case: rather than the small studies' converging on the same result as a large study when published, the small studies give a very different result. ... Small studies are inherently unreliable - larger studies or, better still, multiple studies on the same topic, are more likely to give definitive, accurate results."