WHAT'S that expression, "Lies, damn lies and statistics"?

It came as no surprise at all to me to learn that a million answers on the latest census had been made up.

Isn't most market research? I know when, as a student, I worked for a leading agency, many interviewers did not hide the fact that they made up answers. And those that didn't fabricate made sure that they got the required response ---the one that the company commissioning the survey was after.

For example: Interviewer: "Now Mrs Smith, what brand of floor cleaner do you use?

Mrs Smith: "Flash -- I've always used Flash."

Interviewer: "Have you ever used Fantasy Floor?"

Mrs Smith: "No, never."

Interviewer: "Might you have used it on those occasions when you couldn't get hold of Flash?"

Mrs Smith: "No, I always use Flash, the supermarket stocks it."

This carries on for the best part of an hour until:

Interviewer: "Imagine a situation in which you have spilt a sack of soot in the kitchen, you're having friends round for dinner and you can't get hold of Flash or any other product other than Fantasy Floor. Would you choose it over and above all others?

Mrs Smith: "Yes, I suppose so."

And that's the answer that is recorded.

I often wonder how many extra (and unwanted) ride-on lawn mowers, electronic cat flaps or multi-use foot spas were produced because of the little ticks we placed in boxes.

Market research is all about manipulation -- massaging figures to meet your own ends. Or getting the right sort or people to answer questions on a subject close to their hearts. A poll organised by the Daily Telegraph recently concluded that the British population was equally divided on the subject of fox hunting. Had it been instigated by The Guardian, an entirely different result would probably have appeared.

Recently I was interviewed in my home by a woman from The Office for National Statistics who questioned me about sex and contraception. For a moment I toyed with presenting myself as a wanton temptress with a dozen regular gentlemen callers and a bedside drawer full of Mates. But, dressed as I was in the style of Emily Bishop, I thought it best to stick to the truth and came across as the perfect person to fill the role of the late Mary Whitehouse.

Statistics are unnerving. All that AB, C1, C2, DE stuff -- sticking people in boxes because of what they do to make a living. Of the samples we used back in the 1980s, the DEs and C1s were always the most helpful, that's something I remember.

The "fictionalised" people in the 2001 census included men in their twenties, pre-school children, women over 85, people who do not speak English and people living in privately rented flats. They were apparently invented to help the Government allocate cash where it was most needed.

If something as important as the census has been made up, it leaves you with little faith as to the results of the many other surveys that we are constantly being bombarded with.

I don't know what the information that I divulged about myself will be used for -- probably for someone to giggle over back in the office while officials fiddle about with more important things, like trying to slash the figures for street crime by 'accidentally' muddling the answers up with a survey on whether women find Colin Firth attractive.