I AM peculiarly well placed to discuss the merits of the Durex Global Sex Survey. For I am writing this on a Jordanian boneshaker of a bus with German doctors, a French traveller and an English lawyer. We are driving from Jordan to Israel, and it seems a good time to talk about national attitudes to sex.

Durex does an annual global sex survey. (By global they mean countries prepared to answer such intimate questions, which is not many.) Our survey said: America comes out first with 28.1 minutes per session, up three seconds from last year. Overall, world sex is down 42 seconds to 17.2 minutes - the German average. In contrast to this depreciative trend, Britain is up 0.2 of a minute to 21.1 minutes. What does this tell us? That if you are not copulating for a full 21 minutes or more in Britain then you should feel inadequate? But it's okay to have five minutes longer for your post-coital cigarette in France where the average is 16.1?

Back on the bus, all it seems to prove is the existence of national stereotypes. "Well," says Benois, 25, equaminously, after I urge him that the French are known for their love skills, "'ow long should it be? Long enough for the man so it's okay for the woman." Very philosophical. The English woman, true to form, is sardonic. "Quick can be great. Ten minutes of good, quick sex is better than half an hour of dull, bad sex."

German doctor Matthias is anxious to define terms. "You mean whether the length is directly related to the quality of the experience?"

"Right," I say.

"Which quality is paramount? Is it the length of orgasm or that we should be sensitive to each other?" His heavy emphasis suggests that it should obviously be the latter.

Matthias identifies the crux: what the hell does it matter? Longer is assumed to be better just as is greater frequency. That the more sex you have per week the better. This is quite an assumption. Ask a prostitute.

The benefit of long sex is that there is more chance of some pleasure being derived during it. But equally 28 minutes of bad American sex is worse than 10 minutes of bad Thai sex. (The Thais came in bottom at 10.4 minutes per session.)

The average time for the average woman to go from cold to orgasmic hot is 20 minutes - measured by whom and how I've no idea. The average for men is 10. So maybe the British have got it right in that respect. But apart from that it's down to individual need.

Appropriately this pointless information has coincided with another recent survey revealing our attitudes towards sex surveys.

Apparently we like them because when we hear that 10.3 per cent of people like being covered in grapefruit, it enables us to suggest it to our partner without feeling like a total perv. But science it isn't.

The problem is that people lie. An expert once told me that 75 per cent don't tell the truth in sex surveys - but how could he know?

The only people who answer sex surveys are the kind of people who want to talk about their sex lives. From this unrepresentative sample we are told that the average married couple has sex twice a week. This clearly is not true if you include those with young children and ones who are too busy to answer sex surveys. Those who do answer say what they feel is appropriate, not the truth. The Americans win the duration debate because they like everything bigger and longer. That's why they have the world's biggest obesity problem.

Tips for answering sex surveys:

1. Lie.

2. Attack is the best form of defence. Stand very close to the researcher and say menacingly: "Have you ever thought about becoming a Jehovah's Witness?"