Wired Magazine has an article this month about a philosophy of science guy who followed top scientists around to see what they did with surprising data. Data, that is, that contradicts their hypotheses.
They throw it out.
Our book has talked quite a bit about the importance of experimental confirmation of mathematical conclusions about physics, and it is written by guys willing to push the envelope. Yet, in discussing the idea that Einstein was able to do his paradigm-changing work because he was an outsider (a suggestion made in the Wired article), they caution us that loonies like to use that idea to justify their looniness.
Scientists faced with surprising data, the Wired article says, assume that they've made a mistake and toss the data even though following it up might lead to bold new ideas. They look for problems with the method, errors in their calculations, and bugs in the software. The authors posit that this is because we humans just have a hard time giving up the stuff we believe is true, even in the face of new information.
I think our discussions here show that, at least a little bit. Faced with not-terribly-new ideas from 20th century physics which clash with our direct experience of the world, we're having a bit of a hard time with the cognitive dissonance the clash engenders.
Subscribe to:
Post Comments (Atom)
This comment has been removed by the author.
ReplyDeleteI havent read that article, and couldnt find it freely online anywhere... but just from the high level presented above, I would disagree with the conclusions. I doubt very seriously that data is truly thrown out. I would suspect that it is, instead, set aside and likely even disregarded for the most part. That might seem like a subtle difference but I think it is very important. The difference is that if it is thrown completely out, it is treated as truly worthless data. If it is disregarded, it is treated as simply irrelevant to the current situation. Irrelevant data can be brought back in at a later time if it is then deemed newly relevant.
ReplyDeleteI would actually think it is best practice to disregard data that contradicts your hypothesis. It would take a lot more discipline to stay focused on a single line of reasoning/observation than to be pulled off into tangent directions any time an outcome was not as expected. On top of that, just as the book notes, most science and experimentation done today is with very complex and highly technical equipment. The possibility for miscalculations and mis-calibrations and even pebkac is probably exceptionally high. I would hope the first thing blamed on unexpected data is a flaw in the process. Only after the initial line of experimentation is exhausted and after that miscalculated data is able to be repeated and then predicted should it start to be focused on heavily in my opinion.
Of course this isnt going to be a universal set of practices and I'm sure we've had many great discoveries by people who became intrigued enough with surprising data to begin focusing on it instead.
I would also hope that these "top scientists" are highly trained and experienced in dismissing any initial cognitive dissonance.
Bah I guess we can't edit posts directly. I'll have to be more careful on typos =/
ReplyDeleteWhat you've pointed out is not only an excellent point on its own, but one of the main things to consider about science reporting: you don't get the raw data.
ReplyDeleteI think the thing that interested me is our tendency, as humans, to reject things that don't fit our preconceptions, whatever they may be.
Yes, that tendency fascinates me as well... in may different subjects. =)
ReplyDeleteWith science, I tend to err on the side of idealism over natural human behavior. Part of that is probably because I dont want science tainted with subjectivity (even though it certainly sometimes is). The more realistic part though is that I think science is extremely competitive and painstakingly critiqued. If you dont leave your preconceptions at the door, your colleagues will be right there ready to tell you all about it.