"Stephen Harris"  wrote
in message news:...
> "Guy Hoelzer"  wrote in message
> news:bei63t$1q1l$1{at}darwin.ediacara.org...
> > in article behg8b$1j7s$1{at}darwin.ediacara.org, Peter F at
> > fell_spamtrap_in{at}ozemail.com.au wrote on 7/9/03 9:35 AM:
> >
> > >> in article bdqgil$f41$1{at}darwin.ediacara.org, Stephen Harris at
> > >> stephen.p.harris{at}worldnet.att.net wrote on 6/30/03 4:19 PM:
>  
> > [moderator's note: I could not agree more; this is why the pejorative-
> > seeming use of the term "sampling error" bothers me
here. It is NOT
> > sampling error: it's sampling. A subsample of a variable population
> > will almost always not be perfectly representative; this is not an
> > "error" as if there were some RIGHT way to sample, it's
just how the
> > real world works. - JAH]
> >
> >
> 
> This is also true of the founder effect where the alleles only partially
> represent the parent population. The reason, IMO,  McGinn makes an
> issue of "sampling error" is because he thinks/thought that
it referred to
> miscalculations of statistical data which were performed by some scientist.

Josh's point completely destroy your argument.  What 
he's saying is that sampling is just a statistical way 
of saying measurement.  So the term "sampling error," 
is synonymous with the term, "measurement error."  
Nobody would suggest that measurement error is causal.  
But because of the gambler's fallacy people mistakenly 
believe that sampling error is causal.

Same problem, different words.

Jim
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 7/10/03 4:38:51 PM