In article <3f0363f2_1@mk-nntp-2.news.uk.tiscali.com>, says...
"Frank Dwyer" <fdwyer@citlink.net> wrote in message
news:3F03018C.671D246E@citlink.net...
Papa Jack wrote:
Again, 51% of the women surveyed said the government
should prohibit abortion or limit it to extreme cases, such
as rape, incest, or life-threatening complications.
51% = a little over 500 people. Big deal.
Not the biggest sample, but it does mention the fact that there was only a
3% margin of error i.e. p<0.03, more accurate than necessary - p<0.05
usually suffices for most statistical results. I'm more worried about the
sampling - it doesn't specifically mention how the women were chosen, or if
it was a random sample without any sort of bias. Assuming a reasonable
degree of non-bias in the sample, which is reasonable considering the survey
was carried out by a non-partisan group, then the result is rather
interesting.
It would be interesting to see exactly what the questions were, and
their respective answer rates.
.