Fallout of Faking Data in Science

So cheat your landlord if you can and must, but do not try to shortchange the Muse. It cannot be done. You can’t fake quality any more than you can fake a good meal.
William S. Burroughs

A while ago, Retraction Watch had an interesting article online: “Unusual: HIV vaccine researcher who faked data arrested, faces felony charges“.

I left two comments that illustrate the fallout of the faked data in science, mostly because I — like other commentators — think people who fabricate data should be prosecuted more strongly. And it’s not only about the money they took for which they give nothing back. This fraud is one thing, the real damage is in actually harming (or even killing) humans:

Seems like a perfect case to highlight the dangers of faking data. After all, it might mislead other scientists to focus on something in the hope that it could save lives, yet it is ineffective. Or think a few years down the road and suppose this vaccine would go into human trials, yet it is based on faked data. Unlikely that it ever got this far? So, this researcher was then expecting to get caught? Or did he simply not think about the long-term effects of what happens when you fake data?
Sad that you need such extreme cases to see the damage to science and human lives, but there you go. Screw the money, think about the damage to people.

After a comment about the opportunity costs of the money not spend for something that actually might help find a vaccine for HIV:

I still wonder though how this researcher thought it would go on from his faked data, or whether he thought about it at all. I mean, you can’t stop faking data if there are no effects, and one day it’s applied/tested by others, so it will come out anyway. What do you say then? Sloppy research? An inevitable result of probabilistic science? Or did he think he would find something better and sweep that research under the rug? Or just leave to another topic/greener pastures? Or was it the stubborn hope that this study just failed to show what should be there and would be there another time (like with a type II error)?
Well, anyway, I hope in cases where there’s clear evidence for fraud, there will be more lawsuits. With topics like HIV vaccines perhaps even one day going so far to make a case for manslaughter. Science is messy and results often are far from conclusive, no reason to further obfuscate the issue with faked data. And even if the own part in science is miniscule, there’s still a responsibility. Otherwise, why do science at all?

There is an old saying that when you are under a lot of pressure (under which many scientists are), your view of the world turns into a tunnel view. You only see the immediate obstacle. I’m not sure, but my guess is that something similar happens with scientists who fake data. Or they just don’t care about the topic, or don’t think they have any meaningful impact. But whatever the reason, they are wrong in science. Pressure is not an excuse, unless people are willing to claim that all scientists act this way. And personally, I think they need to be prosecuted much, much more strongly. Not only in areas like HIV vaccine research, but in general.

If science is worth funding, then people apparently care about the issues. So coming back with wrong answers or “snake oil” solutions not only wastes that money. It makes an already relevant problem worse.

2 Comments

  1. It actually makes my blood boil knowing the fact that someone could actually be so pathetic as to lie to the world about something so important. Why? Is a main question I have and, how could you call yourself a scientist when YOU ARE NOTHING BUT A CROOK. Thank you, have a nice day. >:(

  2. Yep, totally second that emotion. Even worse when crooked scientists go on Retraction Watch to debate in favor of themselves under a pseudonym. And yup, that also happens.

1 Trackback / Pingback

  1. Using the Fraud Triangle to Explain Scientific Misconduct | ORGANIZING CREATIVITY

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.