Freakonomics and Divorce
I'm sorry, but Freakonomics is just cool - both the book, which was actually a very interesting read and the website. The tool that is currently interesting me is a widget located here: http://www.divorce360.com/content/divorcecalculator.aspx.
Darren and I are fine :). Friends of ours found this website through a Freakonomics article and it's fun to play with the statistics.
The way it works is based on a series of historical socio-economic data it determines the percentage of people with similar backgrounds who are already divorced and the percentage of people with similar backgrounds who will be divorced.
In my case, 3% of similar people are already divorced and 8% will divorced within the next 5 years. Not so interesting... until you start to play with the levers. So I am female, I have a child, I was 25 when I got married, I have a university degree and I've been married for 7 years.
If I remove the child from the equation, 22% of people were already divorced and 16% of people will be within the next 5 years. So having Lochlan has a significant effect on MY likelihood of staying married. However, children aren't statistically significant for men. At least that's what I assume because the widget doesn't ask the men if they have children.
So for Darren - with his correct data, 8% of men were already divorced and 5% would be within the next 5 years. So the longer we're married the more likley it is that I'll get divorced, but it's less likely that Darren will.
If you reduce my education level, my risk goes to 10% already divorced and 11% will be divorced within 5 years.
Couple of things to note... I don't know how they're gathering this data. I assume it's based on census data, Amercian census data, so it's probably not at all applicable to me because I'm Canadian and Canadian divorce rates are quite a bit lower than US divorce rates. All it's doing is telling you how many people in the US with similar background are divorced or will be divorced.
It doesn't really MEAN anything - it's just a way of reporting census data, but it's interesting to see the effects of these different levers on divorce rates. I'm just a statistics geek. Sorry.
But some of the levers have interesting results when you start to play with then. So if you're bored on a Friday afternoon... play with statistics!!
I thought I had a point... but it's been lost in the busy-ness of the day. But I'm posting this anyway... so ha! This is on long pointless post about a statistics widget I found.
Now I'm going to build a transportation zone system. Fun, fun!