home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!spool.mu.edu!agate!mlloyd
- From: mlloyd@ocf.berkeley.edu (M. Lloyd)
- Newsgroups: soc.bi
- Subject: Re: Introduction, or De-Muffining
- Date: 22 Dec 1992 01:45:54 GMT
- Organization: U.C. Berkeley Open Computing Facility
- Lines: 39
- Message-ID: <1h5s0iINNlbt@agate.berkeley.edu>
- References: <4pp2lv=@rpi.edu> <1gutjlINNi4a@agate.berkeley.edu> <1992Dec21.112536.4176@dcs.warwick.ac.uk>
- NNTP-Posting-Host: earthquake.berkeley.edu
-
- kay@dcs.warwick.ac.uk (Kay Dekker) writes:
- >mlloyd@ocf.berkeley.edu (M. Lloyd) writes:
- >>safarj@rebecca.its.rpi.edu (Jonah Emanuel Safar) writes:
- >>>S.D.: 6 2/3" (For small samples, remember, divide by n-1)
- >> ^^^^^
- >>Splutter?! Remove that word this instant. The only situation where
- >>division by n is appropriate is if you have sampled the whole
- >>population. Now in Joe's case, I might have believed this had happened,
- >>or will soon :-)
- >
- >Why, Mike? where n is large, the difference between n and n-1 is
- >so small that it makes (asymptotically) no difference; at least, that's
- >how my A-level stats teacher explained it.
-
- Well, it is true that 1/sqrt(n) and 1/sqrt(n-1) do get awful close
- togther as n increases, but really, it's not as if it's any difficulty
- to do the mental -- operation as you're bashing the calculator. If
- you're doing it by computer and you use n, you're daft. What are you
- going to do, have an if to isolate "large" n from "small" n? No, n has
- its proper place (the population standard deviation) and it's wrong,
- albeit often close, to apply it to a sample.
-
- Techstuff - ick! Bar the marking, I just *ceased* to be a working
- statistician, and am glad of it. So why am I provoked to post? Well,
- Kay, you did use the "a-word", asymptotically. It's not advisable
- within my hearing. I turn into one of those monothematic bores who
- rabbits on and on and on about the way the New Physics has its head up
- its ass, and they should have clued to this stuff 60 years ago (if not
- sooner). Asymptotics Suck, the world is not deterministic.
-
- It's reality, it's random, get used to it.
-
- hugs
- Mike, who has no doubt of Kay's grasp of reality whatsoever, btw.
- --
- Mike Lloyd, B0/1 h- f- t w- g+ k+ s m- e? | "Bloody nose and burning eyes
- Retro-hippy, music nut, bi and | Raised in laughter to the skies"
- backrubber of devotion | - Bruce Cockburn
- --The end of confusion is the beginning of death--
-