Early on - maybe 1981
Back on 11/13/13 I posted a blog entitled “Beware Greeks
Bearing Manuscripts”, which featured a very famous bio-statistician named Dr.
John Ioannidis. I had written about Dr.
Ioannidis earlier (5/22/12). Dr.
Ioannidis almost singlehandedly has stirred the bio-medical community into a froth. His influence seems to have originated with a
paper he published in 2005, entitled “Why most published research findings are
false.” At scientific meetings I’ll
bet he eats lunch alone.
Dr. Ioannidis seems to turn up everywhere. A recent article in The Economist:
informs us that he is starting an institute at Stanford ,
the purpose of which is to examine bio-medical research and judge how reliable
it is. The name of this menacing
creation is the "Meta-Research Innovation Center at Stanford": METRICS for
short. One thing METRICS will do is to
troll for studies with faulty statistics; that is, studies liable to produce “irreproducible results”. Science is supposed to test the validity of
new discoveries or pronouncements by seeing if the results cited can be duplicated. This process is explicitly performed very rarely, however, for obvious reasons, mainly involving monehy and professional advancement.
The goal of METRICS is to hold comfortable scientific feet to the fire. METRICS also will look for evidence of" publication bias", an evil that is poorly explained in the Economist article, but apparently pretty bad.
The British medical publication Lancet highlights the need for quality control in medical research:
they reckon that $200 billion in annual medical research spending (85% of the
total) is “squandered on studies that were flawed in their design , redundant,
never published, or poorly reported.”
Ioannidis and his METRICS buddies are taking aim at this situation which,
if only partly true, is still outrageous.
Good for them.
Better to fund
institutions like the Rivkin Center, which awards support only to carefully
selected investigators with precisely designed and targeted experiments.
P.S. Oddly enough,
the Greeks and Manuscripts entry has attracted an unusually high number of
hits lately. Why this should be is a mystery to
me.
Really, you should read the Economist article. but because I know you won't, let me quote from Dr.I himself, commenting on METRICS. "I don't want to to take for granted that any type of meta-research is ideal, efficient and nice. I don't want to promise that we can change the world, although this is probably what everybody has to promise to get funded nowadays."
ReplyDeleteInteresting comments on the article. (Didn't you write an article on a similar subject, 'way back when?)
ReplyDeleteYew, indeed I did. I first found out about Dr. I. in an article in the Stanford magazine. However, I searched for an hour and couldn't find it.
DeleteThanks for reading this stuff for all these many months.
Interesting but kind of discouraging. It seems to me that his organization needs to be working with people as they design their studies, rather than evaluating them after the fact when the money has already been "wasted."
ReplyDeleteThe serious among my thousands of readers should really take the time to read the following article:
ReplyDeletehttp://www.theatlantic.com/magazine/print/2010/11/lies-damned-lies-and-medical-science/308269/