This week’s Economist contains an interesting article on a study evaluating the accuracy of published medical research articles. The conclusion:
When Dr Ioannidis ran the numbers through his model, he concluded that even a large, well-designed study with little researcher bias has only an 85% chance of being right. An underpowered, poorly performed drug trial with researcher bias has but a 17% chance of producing true conclusions. Overall, more than half of all published research is probably wrong.
This study is a reminder that there is a difference between the predictive power and accuracy of science as an overall, long-term investigatory method, and the accuracy of any one particular study, particularly in areas such as medicine or biology, where variables are hard to control. To say that one conclusion is the result of “scientific” reasearch does not mean that this one conclusion is necessarily correct. But conclusions tested and found true again and again are probably correct. The devil is in the details. Those who assume unquestionably that because something is labeled “science” it must be true misunderstand what science is supposed to be: a process, not a promise of truth in every instance.
The study itself can be found here. I confess I have not yet read it, only the summary in the Economist.
Categories: Science