survey result: what's true about you

I have done some initial analysis of our silly little survey, friends, and the results are ASTOUNDING. (Actually they’re totally predictable, but I can hardly say that, can I?)

Would you like to know what they are? All righty! (May I have a drumroll please?)

You, dear reader, are, on average, a 31 year old woman who wants sex 5 times per week.

This is where you say, “Well, that’s not true for me.”

And indeed I can state quite definitively that it’s NOT true for you. In fact, no one was 31 years old.

Two of you were 32, and neither of you wanted sex 5 times per week; one of you 32 year old women scored as 3.5 times and week, and the other scored as 18 times a week. And two of you were 30 – one at 7 times per week and one at 3.5.

The point I want to illustrate is this: you can’t determine ANYTHING about an individual from the group average. The average describes is the GROUP and ONLY the group. When you hear an average, you don’t know anything about the individuals in the group.

That’s not precisely true. You could say, “There’s a 90% chance that a person who responded to the survey would ideally like to have sex somewhere between 1 and 10 times per week.” That would be true. Even for the people who want sex more than 10 times a week or fewer than 1 time a week, it’s still true that they had a 90% chance of wanting sex 1 to 10 times per week. Just because you happen to be in the 10% for whom it wasn’t the case doesn’t make you less a part of the group, and 90% of the group wanted it 1-10 times a week.

But have you learned anything interesting or, more to the point, useful, about a person when you say they have a 90% chance of wanting sex 1-10 times per week? No. 1-10 is a gigantic range, and there’s a 10% chance even THAT huge range isn’t right for the individual.

An average – in this case, I’ve reported the median – is not intended to be descriptive of individuals. When people talk about averages, they’re not talking about individuals. They’re describing a group.

I want you all to remember this. I want you all to lock it away in your hearts and carry it with you always. And most of all I want you never again to say, “But that isn’t true for me!” when you see an average. Of course it isn’t true for you. It’s not intended to be true about ANYONE. It’s only intended to be true of the group.

This is one of the few ways in which your responses paralleled national representative samples: the distribution was profoundly skewed. A very small number of people reported wanting a whole, whole lot of sex, while most folks were piled up over on one side. Distributions of numbers of sex partners follow the same pattern.

When the distribution is skewed like this, if I tell you the MEAN, I’d be giving you a falsely high number. The median gives a more representative number.

More representative, I hasten to repeat, of the GROUP. Not of any individual in the group.