Lies, damned lies and statistics

Patron saint of lost causes

There has been some excitement among the anti-smoking ban lobby about a survey among pub licensees that showed 76% thought smoking should be allowed in separate well-ventilated rooms in pubs, and 64% felt it had been detrimental to their business.  The survey was in the Morning Advertiser, the newspaper for the pub trade; the article is here, and I came across it via Curmudgeon's blog.

This looks fairly decisive, until you actually analyse it.  156 licensees responded to the survey, 76% of which is 119.  With 52,000 pubs*, this represents 0.23% of the total number of licensees in the UK; in other words, fewer than 1 in 400 licensees actually responded that they wanted smoking in separate well-ventilated rooms. The sample is so low it is statistically invalid, and I really doubt it was scientifically selected. Like most newspaper surveys, it was probably self selected, and such surveys are notorious in that those with a gripe are most likely to respond, thus skewing the results. While I'm sure there are quite a few licensees who regret the smoking ban, this survey is utterly worthless and proves nothing. The attitude of most licensees I know is, that's the way it is, we just have to get on with it, but that's not a scientific survey either.

In February, I told you about an attitude survey of 1142 students by the National College of Legal Training which, among other things, showed 90% opposed to the repeal of the smoking ban. They're not only the drinkers of the future, but also the next generation of the UK's politicians, lawyers and opinion-formers, so it's reasonable to conclude that's the direction things are moving. People who put faith in the Morning Advertiser survey and its ilk are desperately clutching at straws, especially in the light of my previous posting.

* Source: the British Beer and Pub Association.