Representativeness - the demographics

J Clive Matthews
15 October 2007
Are these people representative?

I've got the breakdown of the demographics of the participants, and need to see just how closely it correlates to the population of the EU as a whole - because, after all, one of the three criteria for success is to see if the sample is "representative", and pretty much the only way to check that is to look at the demographics.

However, I've got a few problems. First, I'm no statistician. Second, all they've provided is a comparison of the actual 362 participants and the initial sample of 3,500, rather than stats from the whole EU (which are notoriously tricky to get hold of). There's also no information about the urban/rural split, incomes, race or religion, which strike me as major oversights.

My feeling is that this was not a representative sample. As already noted, there wasn't a single black face amongst the participants for starters, with only two that I spotted who were non-white. Can anyone tell me the probability of a random sample of 400 people out of nearly 500,000,000 having only two non-whites, when most estimates put the average non-white population of the EU at around 5%, meaning there should have been around 20? And can anyone tell me if this then means that the sample isn't representative, or whether it doesn't really matter?

Based on clothing alone, I'd also guess that the majority were loosely middle class - possibly born out by the figures that 57.7% of the participants have had at least some university education, while only 8.6% didn't finish secondary school. There is, however, no indication of income in the figures released so far.

Plus, of course, there's the whole question - which applies to pretty much any poll, but is still a valid one - of whether people who respond to polls can ever be fully representative. When I've had phone calls asking voting intentions and the like, I always hang up - same with most postal surveys.

I did have a chat about all this with Marc-Andre Allard, one of the organisers from TNS Sofres, and Professor Luskin, and will post up my notes of these conversations later.

I'll also post a link to the demographic breakdown as soon as I can - but if anyone wants to help me number-crunch, leave a comment or get in touch (via james.clivematthews [at] opendemocracy.net) and I'll send you a scan of the information I've currently got. Thanks!

Had enough of ‘alternative facts’? openDemocracy is different Join the conversation: get our weekly email


We encourage anyone to comment, please consult the oD commenting guidelines if you have any questions.
Audio available Bookmark Check Language Close Comments Download Facebook Link Email Newsletter Newsletter Play Print Share Twitter Youtube Search Instagram