Scientific papers are often difficult to understand. The fact that they are often poorly written doesn’t help matters, but many of us lack the scientific background to quickly grasp and analyze scientific literature.
Here is an excerpt from an article published in the British Medical Journal that provides some help reading research papers. Titled "How to read a paper: Assessing the methodological quality of published papers," the article by Trisha Greenhalgh walks the reader through a proper method of determining whether the reader should change his or her practices based on reading a paper.
The key questions:
"First, was the study original?
Second, whom is the study about?
Thirdly, was the design of the study sensible?
Fourthly, was systematic bias avoided or minimised?
Finally, was the study large enough, and continued for long enough, to make the results credible? "
As you might imagine, the article can be used as a tool for developing a cross-examination of an expert who has relied on a questionable study.