Along with some other enlightening speakers, I just had the pleasure of listening to a full day of lecture material by Dr John Ioannidis. Dr Ioannidis is a leader in the field of Epidemiology and Research Statistics. In 2005 he authored an article entitled, “Why Most Published Research Findings are False”.
I’ve read his articles before, but having the man himself run through each of the areas covered in the article was a real eye opener. I don’t want to be critical of all the dedicated and hard-working researchers out there. We owe them a debt of gratitude. But, it would seem that the publishing of research findings is a an area where accidental (or non-accidental) misleading statement, conclusions and assumptions are often made.
The biggest issue we have is the pursuit of findings that are often based on research with a formal statistical significance, commonly with a p value less than 0.05. Dr Ioannidis shows that these findings can often be incorrect. He should know. He’s been involved in large scale studies which made findings that were later refuted and shown to be (more than likely) incorrect.
As practitioners, we are called upon ever day to make decisions based on the published research, so it is important that we have a good understanding of what we are reading. I recommend reading the article by Dr Ioannidis and taking note of the corollaries he mentions. It is this sort of introspection which makes us better practitioners.
Please, read his article. In fact, don’t read it, study it, and get interested in what all those statistics actually mean.
John Baez does a nice job of simplifying the corollaries, and there is some good discussion to follow it on his blog.
The article is free here: http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020124