Does living by the seaside make you ill? Today, we’re showcasing some excellent reporting by Shaun Wooller of the Mail because it’s a perfect example of how to interpret statistics, and more importantly, how not to interpret them. And to strike a personal note: we think the discipline of statistics might give us a way out of the muddle the world has got itself into by relying on emotion, unreason and by inveterately jumping to conclusions. But first to Shaun’s article [1]
He quotes huge, statistically robust studies made by the impeccable University College London. Who find, unarguably, that if you live in one of Britain’s seaside towns you are going to have a 21% higher chance of developing chronic health conditions such as diabetes, cancer and depression. At first this seems counterintuitive: surely all those views and sea air are good for you? What dreadful maritime factor is at play? Should our seaside towns be evacuated at once? Fortunately, Shaun has the answer on hand: look at the samples. If you do that you find that A all the younger people tend to move away from such towns and B so do the affluent. Leaving behind a population of elderly, poorer residents who are, ipso facto, more likely to suffer from these ailments anyway. Which leads us to our own conclusion: sea air may be good for you :but statistics are much, much better
Long standing readers will recall our piece about the dangers of Ice Cream (LSS 23 4 20) in which we covered the irrefutably demonstrable fact that as ice cream sales increase, so does the chance of death by drowning. From which the hysterical and the ignorant might conclude that ice cream represents a clear and present danger to our young. Taking no cognisance of the fact that ice cream sales increase in warm weather, when more people go wild swimming in lakes rivers and oceans, massively increasing the drowning statistics for that week.How about banning lakes and rivers instead?
As Shaun shows us, as Professor Spiegelhalter does even more so in his excellent work {2] the discipline of statistics is a master class in critical thinking. It reminds us that, while many observations produce patterns, finding the cause of these is much more tricky. That the base rate is all important and that every sample made will necessarily be imperfect. Above all that all learning has limits, and what those limits may be. In a world on fire, where warring tribes now clash daily those may turn out to be very important lessons indeed.
[2]Spiegelhalter, David. The Art of Statistics: Learning from Data. London: Pelican (Penguin Books), 2019.
#statistics #health #epidemiology #medicine #survey #sample #cancer #diabetes #poverty #age
Learning, Science and Society 