This just in! A Kansas State researcher once discovered that if you compliment a woman about her appearance, she feels better about herself.

Ooh! And in other news, another researcher (also at Kansas State) determined that familiar songs act as memory prompts; amazingly, hearing a song from days gone by reminds you of a specific time or place.

This must mean that:

1) Academics really are nerdy types that don’t go out on dates very often.
2) Researchers really need to read more marketing how-to books, because advertising wonks discovered the benefits of music for memory association, oh, about 75 years ago.
3) The whole creationism vs. evolution debate in Kansas is already having an effect on the quality of research done there.

Okay, to be fair, Kansas universities aren’t the only places producing … interesting studies. The University of Durham recently made the headlines because researchers there announced that they’d discovered that sports teams that dressed in red have a slight advantage over those that don’t. I’d be willing to bet money that their next study will be: The Effect of Scientific Research on Sports Wagering: Red Teams Get Shorter Odds.

Personally, I’d like to see a study entitled: How the Heck is This Sort of Research Getting Funding Anyway? Or how about: Why on Earth is This Stuff Being Published?

Published research is supposed to be filtered by what’s know as the ‘peer review’ process. If you’re an academic, you’re supposed to do your work, write it up in a paper and submit it to a journal. There, a panel of your peers examines it, and either rejects it, suggests ways to improve it, or gives it the green light for publication.

This little system works well except in the following situations:

1) If you’re doing pioneering research, and there aren’t enough other people in the field for you to actually have peers yet.
2) If your paper argues against a currently accepted theory, in which case you’re liable to get your paper stamped “crackpot” as well as “rejected.”
3) If you have recently insulted, criticized, argued against, or accidentally cut off in traffic, any of the people who end up on your review committee. In this case you’re liable to get your paper stamped “mother dresses funny,” “crackpot,” “should have credentials revoked” as well as “rejected.”

With this kind of process in place, how are some of these studies being published? It may be that there are just too many people submitting too many papers.

I once dated a mathematician who admitted that he wrote (and had published) the same paper several times over: he’d run a math experiment with one set of numbers, and published the results. Then he’d change the input numbers slightly, run the experiment again and published the results. Which just goes to show: academics do go out on dates from time to time.

No, wait, what it really points out is that perhaps journal reviewers are overwhelmed, and aren’t reading papers as closely as they could be.

We know this to be the case on at least two occasions: Alan Sokal submitted a rubbish piece to “Social Text” a few years ago, had it accepted, and then revealed his hoax in another journal. More recently, some computer science jokers wrote a program that generates complete papers for you (http://www.pdos.csail.mit.edu/scigen/). The authors of the program submitted a generated paper to a conference — and had it accepted.

(There’s also the issue of crappy reporting on studies by the media, but that’s a whole other post!)

All of which leads me to ask three more questions:

1) How many other nonsense papers are getting published that really shouldn’t be?
2) Is knowledge really doubling every five years, like the pundits say? (Not really, if the studies above are any indicator of what’s being touted as “knowledge.”)
3) How cool is it that I ‘wrote,’ in less than five seconds, with my eyes closed, a 1500 word paper called “A Synthesis of Forward-Error Correction Using Mir” ?
3b) Where can I get one of these autogenerator thingies for humour posts?