Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site rocksvax.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!princeton!rocksvax!dw
From: dw@rocksvax.UUCP (Don Wegeng)
Newsgroups: net.singles
Subject: Re: Christianity, Statistics, Sex and Marriage
Message-ID: <367@rocksvax.UUCP>
Date: Sun, 26-Aug-84 16:34:08 EDT
Article-I.D.: rocksvax.367
Posted: Sun Aug 26 16:34:08 1984
Date-Received: Mon, 27-Aug-84 00:32:25 EDT
References: <226@uwmacc.UUCP>
Organization: Xerox
Lines: 47

I do not dispute that most readers of Redbook magazine are probably Christians,
for this seems plausible.  However, I will dispute any conclusion that
we can therefore extrapolate the results of a Redbook poll to cover ALL
Christians.  That's plain silly.  I can just as easily claim that
the results of a survey of Republicans who smoke marijuana will accurately
reflect the views of ALL Republicans.  Both might be true, but I cannot
make such a conclusion without more evidence.  Such evidence might
include a study of the corolation between past Redbook survey results
and the results of surveys of Christians in general.

Your reference to the sample size being 100K as another indication
of the validity of your extrapolation is also not reasonable.  The sample
size determines the statistical accuracy of the survey in relation to
the group which is being sampled.  You give us no details about the
total number of Redbook readers who are Christian, much less the total
number of Christian women in the United States or the total number
of both men and women who are Christian.  100K sounds impressive,
but without more details about the population being sampled it's pretty
meaningless.

Finally, it's worth repeating Rsk's observation that the Redbook survey
can only be thought of as a expression of the opinions and beliefs of
people who respond to Redbook surveys.  It's certainly not clear that
this survey even expresses the views of Redbook readers as a whole,
for from what I have read in this newsgroup it appears that this survey
was printed in the magazine, and readers were asked to mail in the
results.  [I must confess that I don't know the details for certain,
and may very well be wrong about this.]  If my assumption is true,
then without further details I can only conclude that the survey
expresses the views of the people who bothered to respond to it.  Previous
studies have shown that people who respond to such surveys often do
so for reasons other than to contribute to the accuracy of the survey.
They often have strong feelings about one or more topics covered; they also
tend to lie [both have arguments have been used to challenge the validity
of several surveys of the salarys that EE's are paid, for example].

In all, I tend to view the Redbook survey results as interesting, and
nothing more.  

-- 
/Don

"Are you mostly engineer, or are you mostly human being?"

arpa: Wegeng.Henr@Xerox.ARPA
uucp: {allegra,princeton,rochester,amd,sunybcs}!rocksvax!dw