XPost: rec.arts.sf.written, rec.arts.sf.science
From: YourName@YourISP.com
In article <9gkEu.281452$PK.1844@fx23.iad>, Chrysi Cat
wrote:
> On 1/23/2014 5:21 PM, Your Name wrote:
> > In article <1xxsbf7bzxt8w.1244utm9kehz9.dlg@40tude.net>, Brian M. Scott
> > wrote:
> >> On Fri, 24 Jan 2014 09:27:46 +1300, Your Name
> >> wrote in
> >> in
> >> rec.arts.sf.written,rec.arts.sf.science,rec.arts.sf.misc:
> >>
> >> [...]
> >>
> >>> The problem is "surveys" are never accurate because they
> >>> simply ask far too few people [...]
> >>
> >> On the contrary, sample size is rarely a problem. The
> >> problem is getting a simple random sample.
> >
> > Few surveys are actually random. Idiot companies like Neilsen pick and
> > choose who can and can't be surveyed in the pretense of obtaining a
> > "statistically equivalent" ratio of people, which in itself means it's
> > never going to be random nor accurate.
> >
> > Sample size is also a big problem. There's no point asking 50 people
> > out of a population of 50 million ... the results are useless, and the
> > way such results are reported is even worse. The survey is ONLY ever
> > going to be accurate for those 50 people. Anything else supposedly
> > "proven" is pure gueswork and manipulation, and often (especially in
> > marketing) statistically manipulated to "prove" whatever the people
> > paying for it want it to "prove".
>
> If I didn't know better, I'd say "Shawn-like typing detected".
> Statistics work. If you couldn't extrapolate from a proper sample (which
> would probably be closer to 500 or 5000 for that 50 mil), then we'd
> pretty much be up a creek for knowing what people do and think...
What do you mean "we'd be"?!? It's already a known fact. Besides which,
most polls and surveys are a worthless waste and piles of money of time
anyway ... the mis-reporting of results of medical surveys is
down-right dangerous.
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
