She Blinded Me With...
If you believe the article, then there is a 50% chance that it's wrong. If you don't believe the article, then the odds that you're wrong go up. I've said it before and I'll say it again...if you want good science, stick with Mr. Wizard.
Posts
7 comments:
This is a particularly relevant caveat when dealing with medical research, for which pharmaceutical cos. shop around (like the tobacco industry before them) for the study that most supports the efficacy of x drug. Cancer research, too, has been muddied for years by cooked numbers, which I know the article you're citing doesn't address. That science can get it wrong as a matter of course is old news, but worth repeating. Was I supposed to say otherwise?
Actually, I just used the article as a vehicle to make my little logic joke. However, I do like pointing out how unreliable the reports on scientific research are. It highlights the reality that science is often as much about generating a buzz and selling books as it is about finding truth. You are correct that science can be manipulated for money buy certain companies, but it is even more often manipulated for acclaim and prestige.
I've maintained that science claims to know more than it does and this article shows one way that it supports that claim.
Nothing really big to add other than, "It highlights the reality that science is often as much about generating a buzz and selling books as it is about finding truth." I didn't get "often" from the data presented. You could say "sometimes" or "can be," but "often" is an assumption every bit as biased as the claims I think you're trying to refute. Talk about biased reporting...
I'd say 50% failure rate would warrent an "often" tag. But then agreeing with the obvious just isn't any fun is it?
What it sounds like you've done with your latest comment is to take the 50% failure rate and attribute that entire rate to "generating a buzz and selling books." Not general errors which may be later corrected. I'm asking what basis your making the distinction?
A major part of publishing for peer review is exactly to locate errors. Bringing in lots of smart people to pick apart your research is a fundamental part of the process. Finding and noting errors is not news. Remember there's a process for theories becoming laws, for example. Jumping in in the middle of the process is just ignorance. Both on the part of the journalists reporting, and the public responding. Generally when you read studies you find words like, "this study finds a statistical link between x and y, in these specific cases." The press and the public then extrapolate from there. Kind of like saying that there are more trees in the US today than when Columbus landed. The original study never even came close to that claim, but its out there, and now ignorant folks repeat it. Then blame faulty science when they're shown that its not true.
Dear Pot,
I will try to more rigorous in my statistical analysis of the articles I comment on.
Love always,
Kettle
Kettle,
Just a reminder, you're part of the media now...
All us pots.
Post a Comment