I’ve received permission from Professor Peter Diggle, the President of the Royal Statistical Society, to publish our recent correspondence. So here it is, minus the start and end felicitations. Here’s what I sent him:
I write to you in your capacity as the current President of the Royal Statistical Society. I am myself merely an English old age pensioner who is becoming increasingly dismayed and bewildered by the seemingly exponentially rising number of health warnings being carried more or less every day in the media.
For example, the BBC yesterday carried a story about processed meat:
Processed meats – such as bacon, sausages and ham – do cause cancer, according to the World Health Organization (WHO).
Its report said 50g of processed meat a day – less than two slices of bacon – increased the chance of developing colorectal cancer by 18%.
Meanwhile, it said red meats were “probably carcinogenic” but there was limited evidence.
And this is just the very latest. A couple of days back there was another story about cheese being as addictive as hard drugs. There have also been innumerable warnings about sugar, salt, saturated fats, and any number of other things, most of which I have been consuming for my entire life, and have no intention to stop consuming now.
My question to you is this: Do you not think that there is a considerable danger that the entire science of statistics is being brought into disrepute by the plethora of these scare stories? I for one never pay any heed to any of them at all. And I don’t know anybody else who does either. In my experience, there is a deepening collapse in public confidence in these sorts of epidemiological studies.
I today visited the RSS website in the faint hope that the institution that speaks on behalf of British statisticians would carry a disclaimer dissociating itself from the media scaremongering currently going on. I watched a number of videos on the website, including a very interesting one about the failure of pollsters to predict the result of the General Election earlier this year. I also watched your address to the society, in the hope that you might address yourself to the growing problem to which I refer.
But I take hope that you may yet actually agree with my concern, given that in your entry in Lancaster University, you cite a pizza restaurant among your favourite eateries, and I very much suspect that you neither believe that you are addicted to mozzarella cheese, nor that you are killing yourself with the salami, bacon, and ham that are often added as toppings to these very tasty dishes.
To which Prof. Diggle replied:
Thank you for your message.
I completely agree with you, and it is not a new problem. Rather it stems from a perennial (and I suspect deliberate) failure of journalists (and others, to be fair) to distinguish between strong evidence of an effect and evidence of a strong effect.
I need no convincing that a big enough sample will show significant differences in a wide range of health outcomes between people who do and do not include processed meat in their diet, but I would expect the sizes of these differences to be very small. As you surmise, I include many things in my diet, and in my life-style more generally (including, for example, driving a car) that increase my risk of death, but not to an extent that worries me.
I do agree that there seems to be a particular spate of these things at the moment. I’m not sure that there is anything new that the RSS could say about it, but I will certainly raise the possibility with colleagues at RSS HQ.
I followed up by thanking him, and telling him I wrote a blog which had just been discussing the RSS. He said he’d take a look at the blog. I may need to write him a further email to explain the nature of the blog and its author and readers.
Anyway, I was delighted to get his email. And after I’d read a little of his colleague’s (Professor Spiegelhalter) blog, I began to wonder if the RSS might be a little haven of reason and sanity in our increasingly demented world.
It reminded me that in the 1950s, Sir Ronald Fisher – perhaps the pre-eminent statistician of his day – was one of the principal critics of the idea that smoking causes lung cancer. He even wrote a book about it: The Cancer Controversy (which is freely available online), published in about 1958. He was also a pipe smoker. He died in 1964. I recently acquired a copy of his daughter’s biography of him.
I began to wonder if the statistics profession had been invaded by upstart non-statisticians like Richard Doll and Bradford Hill, whose disciples now fill the media with scare stories about almost everything, which are repeated over and over again until the general public have absorbed them as the gospel truth as told by anointed ‘experts’.
Might it be, I wondered this morning, that the scepticism of Sir Ronald Fisher still survives in the recesses of today’s Royal Statistical Society, its voice drowned out by the WHO and its baying acolytes in the mainstream media?
But that’s probably just wishful thinking.