Mombu 2008-01-26 12:03:09
Can someone help with the following?
The success of a particular surgical procedure has been described as
being rare, according to an expert in the field who has direct and
indirect access to 10s of 1000s of records.
A small forum has found amongst its members that the success rate is
80% (24 out of 30 respondents to a poll) – very high! ( It’s
acknowledged amongst the forum members, that determination to seek
advice and knowledge about their condition has improved their
prospects although no-one can possibly gauge to what extent – in any
case I want to ignore that for the moment)
As far as I can see, our forum poll was adequately rigorous and I
believe everyone has answered honestly. On the other hand, I suspect
the large study could be highly flawed in its assumptions and that the
counting methods employed could have overlooked large numbers of
Given the enormous disparity between the two results, I’m convinced
that 30 is a large enough sample – assuming it is not drastically
atypical from the large population – to draw this conclusion. Am I
I’d be most grateful for any insight or thoughts on this.
Z 2008-02-29 18:22:45
yeah, 30 is not too small a number, although it is pushing it; as you
suggest, the issue is, whether the 30 is a typical sample of the
population. And, conversely, the bigger study can be worthless even
though it’s big, if its sample is not typical.