"It's a needle in a haystack problem: 99.99% of the data is noise,"
Anonymous
22 Feb 2005 22:33:32 UTC
Topic 12647
(moderation:
> I don't expect to receive any scientific credit. But just posting a list on
> this site somewhere of those that have found pulsars in their WU's would work
> wonders for getting more people involved.
> It would not take much, just something like
> John Doe possible pulsar in WU xxxxx at RA xxx Dec xxx frequency xxx
>
This may not exactly work wonders. I noticed that when I ran the Find-A-Drug program they would list users who's hits went on to be used in labs. Rather than encouraging me I found myself feeling slighted. Logically I knew it was a childish reaction, but I couldn't keep my mind from thinking that my effort was wasted. It was actually discouraging.
I've seen these types of arguments on just about every distributed computing project discussion board there is. DC is just an interesting solution to a common modern problem. The users are basically just philanthropists. If you were one of the thousands who donated money to the American Heart Association and they made a breakthrough discovery that year you wouldn't get your name mentioned anywhere.
The only satisfaction you will get from running DC is a personal one. The chance to donate in a bit more hands on way than just writing a check is actually the main point. I don't care about credit or seeing my name in lights. In fact I've beta tested 4 projects, including this one, and 3 of those completely threw out the months of data & credits we had accumulated when they went live. No tears, because the point was getting a new project on the streets not seeing an imaginary number next to my user name.
"It's a needle in a haystack problem: 99.99% of the da
)
> a negative is just as important as a positive in this search, so already all
> of us have been part of a success. that's the way i look at it anyways, it
> may seem childish but that's how i feel about it.
I've been lurking in this thread, not saying anything, but I just wanted to echo this point. One of the distinguishing characteristics of 'real' research work is failure. The truth is that most (almost all!) 'real' research fails. In fact, if something doesn't fail most of the time, it's not really 'pushing the envelope' and I'd claim that it's not really research. And 'failed' research is really not failure, since it establishes that something doesn't work as one might have hoped. Good researchers make an effort to understand this failure so that they and others can learn from it, and steer a better course in the future.
So the point being made here is a very good one: those people who are NOT finding signal candidates are getting rid of large parts of parameter space and so doing something that is just as important to the collective effort as the positive results of those people whose computers find follow-up candidates.
I've written something about this in the FAQ (front page). I'll add a bit to that in the future.
Cheers,
Bruce