Thursday 28 November 2013

Fun with numbers

From yesterday's Metro:

So, women have only had 7.7 sexual partners on average but men 11.7?

Surely in a population split roughly 50/50 for men and women, the average should be the same. Somebody's telling porkies, ladies.(1)

Junk, London.


From today's Metro:

Junk, have you not considered that men are overestimating their sexual conquests rather than ladies telling fibs?(2)

Roxanne, London


and…

Junk, I agree, women are just as promiscuous as men. I had friends at university with double digits of sexual partners.

However, I'm 32 and have only had two sexual partners.(3) The laws of average don't apply to the whole nation.(4)

Jo, Leicester.


1) I think that the mathematical average (mean) must be the same for both, but it is quite possible that the median values are 7.7 and 11.7.

2) Apparently women count it differently. If they have/had a partner (boyf or husband) with whom they have/had sex, that ticks both boxes "sexual" and "partner" so that counts. But a one-night stand is not a "partnership" so it doesn't count. Men miss this subtle distinction and count everybody.

3) Also, men boast and women are more modest/judgmental. As illustrated by this letter.

4) But the women is clearly an idiot anyway.

11 comments:

Tim Worstall said...

A usual explanation of the difference between male and female numbers of lovers is that some women, ie the village bike, have many many partners. Thus, as you say, mean is roughly equal but median ain't.

Mark Wadsworth said...

TW, yes, that is what I was alluding to, but I don't like the expression "bike".

a) It is insulting to open-minded women and probably women generally, plus

b) they make a lot of men happy, so if we shame them out of existence, the sum total of human happiness goes down.

Anonymous said...

but it is quite possible that the median values are 7.7 and 11.7

No, because that would mean that someone somewhere had .7 of a sexual partner...

Assuming no homosexual pairings and full accurate reporting, then the total number of pairings would be the same for both males and females (by definition) and the means would be proportional to the differences in population size. As the population is fairly evenly split, you would expect similar means. The difference is down to either the sample error (as not *everybody* was asked), and/or the assumptions being false.

Mark Wadsworth said...

F, fair point. The median value ought to be a whole number.

So perhaps they took their sample and did the decent thing of ignoring top and bottom five per cent as being outliers, and then averaged the remainder?

Anonymous said...

I wonder if men have included tarts.

Anonymous said...

MW, it could mean anything at all really.

Mark Wadsworth said...

Anon, that makes no difference to the logic of it.

F, indeed. I suspect that the true mean is 9.7, end of.

Lola said...

Crikey! I have been missing out...must do something about it...

Bayard said...

4) I think what Jo is trying to say is that the sample was unrepresentative of the population as a whole, but didn't know how to say it in so many words and so produced a nonsense.

TW: If you do a quick mental exercise with 100 men and 100 womwne, you will find that the average is always the same both sexes, no matter how many partners each man/woman has. To take your cycling analogy, if twenty of the women had sex with all the men and the other 80 were virgins, the average number of partners for the women is still 20 (20*100/100), as it is for the men, and it is still equal if there are 99 virgins and one woman who had sex with all the men.

Mark Wadsworth said...

L, too late.

B, Jo can f- off. The average is the average is the average.

As the great Swedish statistician said, the average number of legs per person in Sweden is 1.995. Jo can point out that she has 2 legs (or indeed only 1 or 0) until she is blue in the face.

DT said...

Surprised no-one has mentioned the rule of three: men multiply by three (exaggerating) and women divide by three (understating). Apparently.