british social attitudes 2012

[aka Calum Da Jazbo]

new survey report

we are getting rather tougher minded it seems ...

[mercia]

only 3,000 people asked - is that really enough to draw any conclusions (haven't read the questions yet)

[Pabmusic]

only 3,000 people asked - is that really enough to draw any conclusions (haven't read the questions yet)

I suspect it is. I once had a statistician explain this in great detail (it took ages); it all depends on the margin of error you are prepared to live with. There are the some very complex calculations that will determine the sample size you need. Here's the calculation: http://www.tc3.edu/instruct/sbrown/stat/sampsiz.htm I presume it is the same that was explained to me, but I really don't have the slightest idea.

[aka Calum Da Jazbo]

the point in simple terms mercia is that if you kept on asking more and more people the same questions would you get a different distribution of answers .... and the stats tell us that this is very unlikely to be the case .... since there are costs to asking the trade off is finely calculated ... sometimes in the nature of things the survey outcome will be a bit off and other times more accurate than one could expect .... but on average good enough ...