But unfortunately, I think we are all sensible enough to know that that isn't what the mail is saying.

The mail is assuming that we all interpret this as follows:

"A hundred randomly selected Britons are told that in 10 years time, they will be taken and plonked in a tall narrow alleyway which is a mile long and will be chased by your lorry with electrified razor sharp spikes ya de ya. the lorry will cover the distance in exactly 9 mins - your call".

Forty of these people will then choose not to do any (more) exercise in the intervening 10 years and hope that when faced with the actual lorry they'll be able to do it.

Which isn't quite a sharp a distinction.

(But I like the first story)

I am obviously aware of that, and it is precisely my point. The British Heart Foundation have conducted a poll that has established that 60% of people would not do regular exercise in order to increase their lifespan, not indefinitely but by an uncertain amount of time. To present this finding by saying that 60% of people would not exercise if their lives depended on it is simply wrong.

Your common-sense interpretation of the report is right, of course, but my point is that it's happening at the wrong point in the chain. The statisticians should be using common sense in order to draw realistic conclusions from their data, and they should then present those realistic conclusions. Instead, they are misrepresenting their correct data, twisting it into an absolutely absurd factoid, and it is we readers who have to use our knowledge to try to turn the factoid back into a fact. That we are having to do that is evidence of a problem with the statisticians.

And remember that this is just one example. It's only the occasions such as this one, where the truth is quite easy to derive from the lie, that we can notice it. The problem is that statistics are constantly used in this way, often in contexts that make it next to impossible for us even to spot that the headline figure isn't true, let alone to figure out what is. That is precisely why it is vital that people stop accepting these figures at face value. Someone at The Mail should have questioned a BHF spokesman over such an obvious lie, but they just repeated it. If they can't even get into the habit of querying the obvious ones, what hope is there that they'll ever investigate the more subtle convincing ones?

Sorry - just read the article: it IS the Mail that is at fault here.

The first line is not in quotes - therefore it is the journalist talking sh*te, not the statisticians at the BHF.

Nope, it is the BHF.

I think that running for your life down said alley doesn't count as regular exercise. :-)

The tabloids constantily use poll statistics to justify whatever crusade they happen to be on. They never really publish the questions that people are asked or the locations or whatever. In most cases I've seen the sample sizes are miniscule and shouldn't be allowed in a school project let alone in a national newspaper.

Personally I think it is a bit odd that despite there being a new poll in the papers every day, I don't appear to have been included in any of them (the reported ones) and neither does anyone I've ever met. I've been asked my opinion on Dove Moisteriser and I think that is it. Statistically I must be in some sort of minority.

Of course, no one bothers to question statistics. Journos are too feckin lazy to check on anything particularly if it is likely to take the gloss of their story.

Check this one out for instance:
http://news.bbc.co.uk/1/hi/sci/t...ech/ 7006640.stm

"Arctic sea ice shrank to the smallest area on record this year, US scientists have confirmed."

What is omitted is that the scientists only started recording the ice extent in 1978. 30 years worth of data in the context of the entire history of the planet is hardly worth mentioning. But the reporter uses the data and stats to sensationlise.