I was staring out of a window while eating breakfast this morning at the University cafeteria and was considering statistics I've heard in my life. One of which is funny, the other is a scary kind of funny.
First the funny.
I was reading the newspaper back in Canada a long while back and found a survey column and the question was:
How many relationships have you had?
Men -- 7
Women -- 4
Now one would have thought that perhaps men were more active in the relationship department. But if you think about the data, something is seriously wrong (for staticians, the sample size was 1000).
Let's do the math. Suppose you had a world in which there were 3 guys and 3 girls. Suppose each guy dated the same girl.
The average number of relationships for the guys is 1. For the girls, it's (3 + 0 + 0)/3, which is also 1.
Right? We should expect that the number of relationships between male and females to be about equal. What might cause this discrepancy? Lots of gay men. There is another possibility but the gay men one is funnier.
Next the scary kind of funny.
It is oft quoted that 50% of all marrages end in divorce. As a result of this, most of us would think that we've got a coin-toss change of making it all the way to the end with a partner. Let's suppose that this 50% statistic is true.
Suppose that every married person gets divorced and remarries. Need I say more?
"There are three kinds of lies: lies, damned lies and statistics."
-- Mark Twain
4 comments:
It is oft quoted that 50% of all marrages end in divorce. As a result of this, most of us would think that we've got a coin-toss change of making it all the way to the end with a partner. Let's suppose that this 50% statistic is true.
This is the wrong way to attack this particular claim.
Instead, think of the frequency of marriage versus divorce.
If you get married when you are 35, the life expectancy is about 40 years assuming no divorce.
If you do get divorced, chances are it will happen in the first five years.
So what an interest statistic would be is the time-weighted divorce rate of people.
For example, if your world consisted of two couples, one couple being married for 45 years, and one couple being married for 5 years and then divorced, then the rate would be 10%.
I sort of missed the meaning of "rate" there. What is a percentage/unit time? It sounds like an exponential problem (ie increases 10%/year or something). Mind explaining that in a little more detail.
I think the interesting statistic would be of all first marriages, what is the distribution in years of divorce (of those that do happen).
In the poorly defined statistic of 50% of all marriages ending in divorce, I just came up with an interesting situation that could lead to that 50% answer.
There's other explanations besides gay men...
1. These numbers are self-reported, which means one can always lie (although the motivation for doing so is unclear)
2. Men date more than one woman at a time
3. Men may have a lower threshold for they would consider a relationship
In summary, the cause seems to be dumb, promiscuous men =)
BTW, I met w/Ed over Thanksgiving. He's having a good time in DC.
P.S. What was up w/that party with all the girls? Advise them to come over to Boston =)
Yuti, even is men dated multiple people at the same time, the math would still work out.
Well, if you can find me cheap plane tickets to Boston from Japan I'll recommend them to check out the place and you could be their guide :)
Post a Comment