Wednesday, May 21, 2008

Advertising and the expression of strange theories dressed up as research and science

Advertising and the expression of strange theories dressed up as research and science


Mathematical lies, statistics, and scaremongering

They amount to much the same thing, but before I dive into an analysis of this, illustrated by a current example, let me give you my favourite quote regarding statistics:

Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital. ~Aaron Levenstein

Indeed, and a truer and more concise quote on statistics you won’t find anywhere. Believe me, I’ve tried, and I always come back to Levenstein’s as the best description in under 20 words. There are some great anonymous quotes about statistics but by being anonymous they lack credence. And just as importantly they lack the sting that comes with quotes from famous people. I mean, if you relate an Oscar Wilde quote to someone and say you don’t know who it’s by, it doesn’t have the sting and panache of delivering the quote then rounding off grandly by simply saying two words –Oscar Wilde.

A lengthier but equally dismissive quote regarding statistics comes from one of my favourite authors:
While the individual man is an insoluble puzzle, in the aggregate he becomes a mathematical certainty. You can, for example, never foretell what any one man will be up to, but you can say with precision what an average number will be up to. Individuals vary, but percentages remain constant. So says the statistician. ~Arthur Conan Doyle

Anyway, I am often amazed how various companies; especially motor vehicle manufacturers, pharmaceutical companies, security companies, medical companies, and even individual scientists, doctors, politicians, and bureaucrats use statistics to enforce their claims and to hide the truths contrary to their claims.

Sometimes I get the feeling that people using statistics think they have invoked some kind of sacred shield that prevents questioning or rebuttal and that makes everyone else ignorant to what the statistics in question are concealing.

Here is one so mind-numbingly unconvincing that I choked on my breakfast back on 27th March when it appeared in a small article in my local daily newspaper VF (Värmlands Folkblad), here in Sweden.

The article was entitled (all following translations are my own, which as a professional translator didn’t prove difficult) :

“A roundish waist at 40 increases the risk for dementia”

So, my first thought is, well, I’m in the risk zone then, I’m 46 now, I’ve got a roundish belly and have had for several years, so I’m interested, I’ll keep reading.

So having got past the scaremongering of the title and the first paragraph telling people in their 40s with round bellies that they are going to suffer from senile dementia in later years and espousing the unreliable findings of this research and therefore this strange theory, I got down to the cold hard statistical lies….sorry I mean…facts.

Of course, by now, even before I’ve told you, you know this is going to be about a study in the US!

Well, the study took 6,583 healthy people in the USA and measured their waists when they were aged between 40 and 45.

Thirty six years later, 16 percent of them were found to be suffering from dementia…… OH MY GOD THAT’S TERRIBLE! SIXTEEN PERCENT…OH NO…THAT’S AWFUL!

I better start dieting, I better exercise more ………HANG ON, HANG ON, HANG ON A SECOND……hold your horses, not so fast there…. 16 percent? 16 percent? That’s not even one quarter!
That means that 84 percent …read that again, EIGHTY FOUR PERCENT didn’t develop dementia!

Please now read Levenstein’s quote again: Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital.

So what was the point of this article other than to scare people approaching 40 into exercising and dieting? Well, as far as I can see there was no point because no companies were mentioned, no medical institutions in the US were named, and no individual scientists or medicos were named.

The only answer I can come up with is that once again, someone at the news agency, in this case TT, has been blinded by statistics and thought this particular statistic newsworthy, even though it is clearly unreliable, and that there simply is no news story there.

This kind of reporting of statistics could go on ad infinitum were it allowed to occur on a daily basis, for example, we would end up with reports like this : 88 percent of all Irishmen living in Sweden are married to Swedish women – totally uninteresting, totally irrelevant, totally pointless… the statistic that is, not the marriage to the Swedish woman!

And then someone at VF has also been impressed by this non-newsworthy, unreliable scaremongering statistic and thought, “WOW we must print this!”

Which all goes to show that statistics are simply numbers, instead of words, telling you what someone else thinks, and what someone else wants you to hear, wants you to believe and buy into, and ultimately to buy their product or service or follow their advice.

Numbers, and especially when used as statistics, seem to be, for many people, more trustworthy than words, and many people believe that numbers don’t and simply can’t lie.

Well just like words, it is what the numbers don’t tell you that you need to look into, because every single person who uses statistics has an agenda, so don’t be scare mongered or forced into thinking their agenda is the same as yours, especially when the statistics come from what many view as reliable sources i.e. government officials and medical experts.

Let me leave you then with a quote which perfectly captures the essence of number manipulation:

Torture numbers, and they'll confess to anything. ~Gregg Easterbrook

No comments: