|
Post by Slartucker on Sept 27, 2007 6:49:21 GMT -5
This is patently false. What you mean is that, if I am trying to get a representative sample, narrow intervals will have more noise than wide intervals. The converse of course is that wide intervals will have less information than narrow intervals. And you are certainly right that AFTER collecting data you can merge intervals try to account for noise. There is nothing to be gained however by reducing the amount of information you collect without increasing the potential to deal with noise -- and that's exactly what wider intervals would cause. Less information, equal noise.
I'm not even sure how much "noise" really applies to this, given that it's not a representative sample or a candidate for equation modeling, but simply a list of ages.
Note also that the "exact numbers" you are talking about are in fact intervals. "16" above covers from 16.0 to 16.99... it's an interval of a year.
|
|
|
Post by BioLogIn on Sept 27, 2007 8:18:43 GMT -5
25.94 Meaning I'll be 26 this October.
|
|
|
Post by dni on Sept 27, 2007 9:42:32 GMT -5
2 Slartucker: Ok, ok... Let's imagine that this is 'raw' data without preparation for anything So my datum is '31', rounded down On 'noise' i meant probability of fact that count of real person will be stable time-for-time. If you use intervals, you can have situation near stable without assumptions, for the years (one goes out the inverval, another comes). If you use exact numbers, you have constantly moving dataline with no stable values...
|
|
|
Post by Rycchus on Sept 27, 2007 12:06:18 GMT -5
Yeah. I'm only 20 but I'm getting older much faster than some of the 30-year-olds. That'll totally screw the data!
|
|
|
Post by Dubber on Sept 27, 2007 14:03:53 GMT -5
Yeah. I'm only 20 but I'm getting older much faster than some of the 30-year-olds. That'll totally screw the data! Heh. Try being 40, you half my age whippersnapper. You may numerically age "twice as fast" as me, but my corporeal container will age faster and faster the further I slide over the hill <edit> Though how that is in my best interest in the argument, I have no idea.</edit>
|
|
|
Post by Rycchus on Sept 27, 2007 14:05:51 GMT -5
I know. I can't even drink in your country for another few months. It's great! I love being my age ;D
|
|
|
Post by spez on Oct 9, 2007 7:11:57 GMT -5
Let's say 24 (i was 23 at the time of posting of this poll, but i'll become 24 in a 2 weeks, so...)
|
|
|
Post by vilhazarog on Oct 10, 2007 20:07:17 GMT -5
Well Dubber, I'm up there with you, 39 at the end of the month. That's right, you young whippersnappers, I was playing Space Invaders in the arcade when it first came out! I was thrilled when we got an Apple ][ and I loaded that 4k or so game up via cassette tape and it only took about 5 minutes or so! That's right, games came on cassette tapes and you plugged your computer into your tape player and hit play and loaded that sucker up! However, I've never stopped playing them... playing Neverwinter 2 and Battlefield 2 at the moment, will get the Orange Box (Half life/Team Fortress 2) next, and a friend just today let me borrow his PS3... thinking of buying one. I'll still be playing when I'm 103. Yes, that old man that just took out your 3 defenders and capped will be me!
|
|
|
Post by Rycchus on Oct 10, 2007 21:13:01 GMT -5
Yeah. I'm only 20 but I'm getting older much faster than some of the 30-year-olds. That'll totally screw the data! Heh. Try being 40, you half my age whippersnapper. You may numerically age "twice as fast" as me, but my corporeal container will age faster and faster the further I slide over the hill <edit> Though how that is in my best interest in the argument, I have no idea.</edit> Also I was kind of being sarcastic about the aging faster thing. It was when someone else said that the data as it was was useless because people are getting older
|
|
|
Post by nawglan on Oct 11, 2007 8:43:04 GMT -5
I was thrilled when we got an Apple ][ and I loaded that 4k or so game up via cassette tape and it only took about 5 minutes or so! That's right, games came on cassette tapes and you plugged your computer into your tape player and hit play and loaded that sucker up! I still have my Apple ][ (and a couple Apple ][gs, 3 or 4 Apple ][c, and an Apple ][ with serial number < 1500). Sucked BIG time when you loaded up your tape, waited through the hiss / squeal (the tape player had a volume dial? really?), only to find out you were one number off on the tape counter, or that the tape counter on your friend's tape deck was a bit off from yours... Takes a while to hunt-n-search around on a cassette tape. Loved the 1 meg of ram I had in my Apple ][e. Shoulda seen all the addons I had for Appleworks. The 5 meg external SCSI-1 hard drive was sweet too.
|
|
|
Post by dni on Oct 12, 2007 1:56:32 GMT -5
2 Rycchus:
So if you will use right data intervals, you'll get that some people coming into interval, and some of them going out. Adjusting of interval bounds let you get somewhat of stable statistics.
2 nawglan: Ah, oldies! I love oldies!.. Scratching sound of tape loaders...
|
|
|
Post by Slartucker on Oct 12, 2007 6:31:45 GMT -5
Wait, I thought you wanted us to not use intervals?
|
|
|
Post by Rycchus on Oct 12, 2007 6:57:47 GMT -5
2 Rycchus: So if you will use right data intervals, you'll get that some people coming into interval, and some of them going out. Adjusting of interval bounds let you get somewhat of stable statistics. As you will with any intervals, including the ones Slarty's already used!
|
|
|
Post by dni on Oct 14, 2007 9:04:04 GMT -5
2 Slartucker: On the contrary, i offer to use intervals in poll (i impulsively said 'bad poll')
|
|
|
Post by Kaito on May 9, 2008 13:21:50 GMT -5
17; 18 in August 2008
|
|