Comment "Peer-reviewed" doesn't mean confirmed (Score 1) 1165
I keep seeing things like the above ("I personally only want to see peer reviewed data, nothing else matters.") that lead me to think that people really misunderstand the whole peer review process and what it means.
When you submit a paper for publication, the journal's editor typically sends the paper out to a relatively small group of other scientists for the purpose of answering the question "Should I publish this paper in the journal". The scientists are typically from the field, and usually have published in the journal previously -- i.e., they have a pretty good sense about the kind of papers that "fit" in the journal. The reviewers usually send back the article with lots of comments about how well the article is (or isn't) written, including grammar errors, other articles which the author should consider, critiques of the methodology or analysis, etc. They might even suggest that another journal would be a better place for the article to be published -- and explain why.
One thing that those anonymous reviewers DON'T EVER DO, however, is re-run the experiment/analysis with the goal of confirming or validating the conclusions of the paper. That is done *after* the paper is published, when another scientist reads the paper and says to himself either "Hey, I can use that!!" or "Hey, that's ridiculous!!" and attempts to confirm the results. If he can't, he gets to write his own paper (which is *also* usually peer-reviewed) that explains what is wrong with the original article.
So "peer reviewed" doesn't mean "true" or "confirmed" -- it really just means that the editor of a journal that requires peer review thought that the article was worth publishing. A "peer reviewed" article might be nothing more than easily disproved propaganda if an editor decides to take an advocacy position, has an axe to grind, or is an idiot...
When you submit a paper for publication, the journal's editor typically sends the paper out to a relatively small group of other scientists for the purpose of answering the question "Should I publish this paper in the journal". The scientists are typically from the field, and usually have published in the journal previously -- i.e., they have a pretty good sense about the kind of papers that "fit" in the journal. The reviewers usually send back the article with lots of comments about how well the article is (or isn't) written, including grammar errors, other articles which the author should consider, critiques of the methodology or analysis, etc. They might even suggest that another journal would be a better place for the article to be published -- and explain why.
One thing that those anonymous reviewers DON'T EVER DO, however, is re-run the experiment/analysis with the goal of confirming or validating the conclusions of the paper. That is done *after* the paper is published, when another scientist reads the paper and says to himself either "Hey, I can use that!!" or "Hey, that's ridiculous!!" and attempts to confirm the results. If he can't, he gets to write his own paper (which is *also* usually peer-reviewed) that explains what is wrong with the original article.
So "peer reviewed" doesn't mean "true" or "confirmed" -- it really just means that the editor of a journal that requires peer review thought that the article was worth publishing. A "peer reviewed" article might be nothing more than easily disproved propaganda if an editor decides to take an advocacy position, has an axe to grind, or is an idiot...
