Map of life expectancy at birth from Global Education Project.

Friday, January 07, 2011

Fraud

Scientific fraud is not terribly rare in one sense. Just about every month NIH publishes official Findings of Scientific Misconduct. But these are almost always a post-doc or a junior faculty member who faked a slide in a proposal or something on that level. I'm sure a lot of fudged data finds its way into the literature, and I know for sure that quite a bit of data mining goes on.

For the uninitiated, that's when you look around in a data set for significant associations, and then pretend you hypothesized them when actually, you just stumbled on them. That's bad because it means your tests of statistical significance are invalid and your findings may well be spurious. It's one of the big reasons why a lot of findings don't hold up actually, because they weren't really legit in the first place.

What you're supposed to do when you come across an unexpected association is treat it as a hypothesis and then try to confirm it with fresh data. But the reason why people don't always do that is the same reason the post-doc faked the slides and data gets fudged -- there's tremendous pressure to publish and get funding, and if you don't, you are likely to get squeezed out of academia entirely and you're selling insurance or something. At best your grand ambitions go unfulfilled and you're working as a drudge in somebody else's lab or teaching as an adjunct for a few thousand bucks a course until you've finally had enough. So the temptations are strong.

But the vast majority of fraudulent or questionable publications are basically inconsequential. They don't lead anywhere, and the worst that happens is that better work gets crowded out of the journal and possibly some other investigators gets sent briefly down a blind alley. What is actually rare is a consequential fraud like Andrew Wakefield's that convinces a lot of people of something that isn't true and is actually important. Even Marc Hauser, if he did indeed commit fraud rather than succumb to wishful thinking (a distinction without a difference, I think, since appropriate methods would have precluded the latter), didn't do much substantive harm. The cognitive abilities of macaques are not a matter of life and death.

But for someone to perpetrate a fraud as consequential as Wakefield's takes a very special kind of personality. He wasn't just fudging some numbers to win a grant or get a paper published in an obscure journal on an arcane subject. He went right to the heart of medical practice and attacked what has been, without question, the single greatest boon to humanity ever delivered by biomedicine. (I'll assign clean drinking water and sanitary child birth to public health, since they aren't medical interventions but rather harm avoidance based on understanding of pathogens.)

The only reason an anti-vaccination movement can exist is that vaccination has been so successful. If kids were dying of smallpox and polio and measles today, believe me, Jenny McCarthy would be screaming and yelling to make sure every kid got every shot. People today just have no idea of what human existence was like before the development of modern vaccines. Life was cheap. People didn't expect most of their children to become adults, they just hoped for it. If you didn't die, you might be disfigured by smallpox or paralyzed by polio or sterilized by mumps. Women who had rubella during pregnancy had profoundly retarded infants, who filled institutions. That's how it was.

If Andrew Wakefield had his way, that's how it would be today. And why did he do it? The way I read Brian Deer's story, he did it for money.

2 comments:

C. Corax said...

I sent an email linking to the BMJ articles to a friend. She wrote back and said that a neighbor who hadn't vaccinated her children had to take one of the kids to the hospital. The kid had two different rashes and a high fever. It took the staff at the hospital awhile to figure out that it was measles, because they hadn't seen it in so long.

And so it starts. That was in CT, by the way.

Cervantes said...

That's a sad tale indeed. Kids usually get over measles, but sometimes they have brain damage, among other unpleasant consequences