kyser_soze said:
Well, leaving aside the arguments on the other thread about sample size and whom the original person questioned passed ICM's interviewers along to, it's a piece of piss to find out who and who isn't Muslim - there are several national marketing surveys which ask about religion (TGI being the most well known) so while these are self-entitled Muslims. What do you mean by 'visible clusters' as well?
And outside of the bit about Sharia (which was a biased question), the survey showed that Muslim's in the UK are pretty much of the same mind as non-Muslims about most political issues.
Finding out yes. But where will you find the muslims - that's the point.
For a survey to be valid, you have to go through a certain method.
First you need to esablish what is the
frame which is the people or things you're surveying. In this case the frame is 'all muslims in Britain'.
Then, for the survey to be valid, you need to have a method of sampling that means every member of the frame has an equal chance of being surveyed, or if some don't, the bias has no effect on what you're measuring.
In this case, you need a survey which means that the muslim family who run a takeaway in Lyme regis have the same chance of being surveyed as Mr Miah from 1 the high street, Bradford.
To do that you need a list, with all the uk's muslims, to select your 400 from randomly. Is there such a list? How many muslims live scattered the length and bredth of the uk, living relatively integrated lives? If they don't figure in the sample, then it's biased against what could possibly be a liberal wing of opinion.
By clusters, I mean that you could do the survey by finding (say) 100 Muslims each in Birmingham, east London, bradford and Burnley. That would NOT be a random sample of the frame, but a sample of Muslims who live in those innercity areas, with far more of their own community around them. Or you could do it outside Mosques - again not all Muslims go to mosque.