home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: misc.kids
- Path: sparky!uunet!spool.mu.edu!sdd.hp.com!usc!news.service.uci.edu!ttinews!metis!reid
- From: reid@metis.tti.com (Reid Kneeland)
- Subject: Re: Here a study, there a study....
- Message-ID: <1992Dec21.183940.20236@ttinews.tti.com>
- Sender: usenet@ttinews.tti.com (Usenet Admin)
- Nntp-Posting-Host: metis.tti.com
- Organization: Transaction Technology Inc.
- References: <1992Dec16.115140.119671@marshall.wvnet.edu>
- Date: Mon, 21 Dec 1992 18:39:40 GMT
- Lines: 48
-
- In article <1992Dec16.115140.119671@marshall.wvnet.edu> swk004@marshall.wvnet.edu writes:
- >...infant walkers. Studies on ear infections. Studies on ear tubes.
- >Studies on the rate of c-sections. Here a study, there a study,
- >everywhere a study, study. (Everybody join in now, "ol' misc.kids
- >had a study, eieio.....)
-
- God forbid we should deal in empirical data instead of anecdotes.
-
- >But more interesting, we've got people saying "no, that study is
- >wrong, the researchers failed to take into account that..."
-
- Of course. The methodology in any study must be scrutinzed
- carefully before one can be sure of the conclusions.
-
- >Ya'll are making my point for me. A study is only authoritative
- >until the next one comes out. Seems to me we ought to be using a
- >whole lot more common sense about these things.
-
- Wrong conclusion, I think.
-
- Studies, and statistical inference in general, are the foundation of
- modern science. Done properly, they are the only way that general
- conclusions can be draw from available data. While no one study is
- definitive, one can draw meaningful conclusions by looking at all the
- studies, looking critically at the methodologies involved, and looking
- at patterns which emerge. Saying that X is or isn't true because a
- particular study said so is not valid reasoning if most other valid
- studies have indicated otherwise. But it is just as wrong to throw out
- the whole idea of studies and statistics because of apparent
- contradictions. There will always be evidence on both sides of any
- medical issue; the question is where the *preponderance* of evidence from
- well-constructed studies lies.
-
- Common sense is great, but it is often wrong. Common sense would tell
- us that the Sun goes around the Earth, that heavier objects fall faster
- than light ones, that the chances of a coin coming up heads increase
- the more times it comes up tails, that turning a thermostat way up will
- heat a house faster. But these are all incorrect. There is a place
- for common sense and a place for science. Adn the biggest problem with
- common sense is that it isn't all that common.
-
- =====================================================================
- Reid Kneeland
- reid@tti.com (that's MISTER reid@tti.com to you!)
- Transaction Technology Inc., Santa Monica, CA, USA (310) 450-9111 x2499
- The opinions expressed above do not necessarily etc etc...
-
- Never trust a man who can count to 1,023 on his fingers.
-