Posted by rbrands on January 16, 2010
Just a reminder that Chancewiki has lots of clinical epidemiology stuff. Have you hit a boring patch of 513? Got a computer in front of you? Head over to http://chance.dartmouth.edu/chancewiki/index.php/Main_Page!
Take a look at Chance News 58, where they talk about negative data. And reverse confounding.
Posted in Bias, Chance, EBM | Tagged: Bias, Chance, EBM | Comments Off
Posted by rbrands on January 15, 2010
In short, yes. Can they be biased? Yes. Can they be unbiased? Yes. Kind of like everything else. An article in the New York Times looks at some of the issues when looking at a recent meta-analysis of antidepressants.
Posted in Bias, Therapy | Tagged: Bias, Therapy | Comments Off
Posted by rbrands on January 14, 2010
This article from the Atlantic provides an overview of the ideas I mentioned in class about overall effectiveness of influenza vaccination. They call the membership bias “healthy user effect.” It gives enough information for you to be able to hunt down original sources if you care. Before, this gets controversial, let me say that I got both my flu shots this year. But then I’m one of those people (albeit only a “semi-healthy user” ) who might have produced the bias, right?
Posted in Bias, Therapy | Tagged: Bias, Therapy | Comments Off
Posted by rbrands on December 15, 2009
A reminder to check out chancenews at ChanceWiki. Do their puzzles. Think about chance and bias. Most of the ways that research goes wrong involves macro-, not micro-epidemiology.
chancenews is at http://chance.dartmouth.edu/chancewiki/index.php/Main_Page.
Posted in Bias, Chance | Tagged: Bias, Chance, EBM | Comments Off
