It was the appearance of this in the local press that left me wondering whether statisticians have become disembodied heads.
It was the appearance of this in the local press that left me wondering whether statisticians have become disembodied heads.
Whilst Coco has every sympathy for those who suffer from the disease of the body which we know as diabetes, Coco has little sympathy for those statisticians who though they are irrefragable in the use of the art to which they are devoted, show little or perhaps even no common sense.
The reporter who brought this matter to his attention does however at least stand one step of contempt above the reporter who recently published an article about the discovery of a decapitated head. Well, Coco has never seen nor even heard of such a thing, have you? Decapitated bodies are not uncommon, as are perhaps dismembered ones, but decapitated heads? It seems as unlikely as a dismembered arm. Perhaps the real intention was to say scalped head, you might ask, but no, the find was of a head, a whole head. One can however understand the dilemma of the reporter, this was clearly a head alone, not attached to any body, but to call it a disembodied head would hardly have conveyed the right impression, and in any event, a disembodied something would be rather difficult to find, given that that which is disembodied no longer has a material presence in this universe. Perhaps they really meant decapacitated, but shied away from a word which may only have reflected back upon themselves.
Coco was unable to find a complete copy of the article online, but you may read the article for yourself here:
One can hardly blame the reporter, poor chap, what does a reporter know about statistics? That question is rhetorical, in case it needs to be said. Coco is not unaware that there may be reporters who have had a good, and far better than Coco’s, grounding in such. But one is left wondering whether or not common sense has been applied.
It does not take much effort to see that the statistical result of two and a half times more likely is incorrect, whatever the results of the statistical analysis might yield. Coco would like to suggest that the answer is really much more like one hundred percent more likely, which is to say almost certainly going to be the case, if not actually in reality, without even the need to apply any statistical analysis at all. For it seems so very clear to Coco that although most young people die old, and many young people have died young, it is impossible that anyone who is old shall die young. Old people always die old. So then, let us read again what the article suggested: The risk of early death was 2,5 times greater for those diagnosed before 40 compared with those diagnosed after 60.
That is a wonderfully incredible statement. Please allow Coco to break it apart. In order to be able to say that one thing is greater than another we need to know, if not the absolute magnitude of the two things, the relative magnitude. In this case we are talking about a comparison of the level of risk. So we should have some idea of the level of risk faced by the two gorups, those who are under forty and those who are over sixty. Now perhaps first we should simplify the examination by removing from it the complicating factor of type 2 diabetes, and we shall adjust, if necessary, our findings later for that omission. We shall add a more general comment about mortality also.
So then we must determine what is the risk of dying young for a forty year old and for a sixty year old? Now Coco is not skilled in the arts of the actuary, but it does not seem unlikely to think that although a forty year old may die young, a sixty year old will never die young. Though if someone thinks that a sixty year old may die young, then it is likely that they have discovered the elixir of life for which the alchemists of yore vainly sought for centuries. Perhaps they would care to share their secret with Coco, or at least publish their results subject to peer review and set up production and marketing companies for the benefit of wider mankind. Oh, Coco apologises, you may have noticed, the mention of statistics does rather cause Coco to stray and indulge in flights of fantasy, unlike Leonardo of course who was the precursor of our own* Wright brothers..
So then, back to the point, for the sake of clarity and being able to do some calculations, without suggesting that the numbers are correct, let us say that the risk of dying young at age forty is 1% of 4‰. We have already said that there is no risk that a person diagnosed at age sixty will die young, such a thing is impossible, therefore the risk of dying young for this group is 0% of 4‰. We may then restate the point in the article as The risk of early death was 1%/0% times greater for those diagnosed before 40 compared with those diagnosed after 60.
Now all we need is a mathematician who can tell us how we can reduce the ratio 1:0 to a number that we shall understand. Coco thinks that it means this: The risk of early death was infinitely greater for those diagnosed before 40 compared with those diagnosed after 60.
This can be put in an even more blunt manner that the research suggested that it was certain that those who were diagnosed before 40 would succumb to an early death, whereas those who were diagnosed after age sixty would die, as expected, old. Coco remains unconvinced of the veracity of this argument, as it is clear that not all who are disagnosed before forty die young, some will die old. However, the mathematics suggests it, and incidently it would make no difference what the actual level of risk was for a forty year old (and we therefore do not need to enter a correction for the diagnosis or otherwise of the condition of type 2 diabetes), the result would always be certain, just as the statistics suggests that the ratio is 2.4.
Coco suggests that one should perhaps also note on the matter of increasing mortality with age, that if at age sixty you are discovered to have a condition which reduced life expectancy to about twenty years then at sixty when life expectancy is less than twenty years, it is perhaps of no great concern. You are likely to have been taken before the condition takes you. If however you are discovered to have this condition when you are only forty years of age, when your life expectancy may have been up to forty years, then you would view the condition in a different way and perhaps feel more threatened by it. The discovery of the condition for a forty year old increases the expectation of mortality significantly, but in a sixty year old hardly at all.
Given the uncertainty that this all provides, should we not be glad that there is one who knows all things and declares to us: ‘I am the resurrection and the life. He who believes in me, though he die shall live’ John 11:25. Those who trust the Lord Jesus Christ need not fear. Diabetes may shorten your life on this earth now, but he will raise us up again at the last day to live on a new earth in new bodies where there shall be no more death, from any cause, nor crying, nor tears, but all shall live in love and harmony with him and with one another.
*Much like Formosa, we are waiting for the colonies on the mainland of the American continent to acknowledge their lawful and ancestral ruler.
No doubt the erudite and skilled shall find much worthy of contempt in this the article of Coco, but Coco would wish to remind such that erudition, skill and facts should not be allowed to spoil a good story as long as it is properly recognised that it is indeed a good story.