August 09, 2010
Dept. of Overcoming Confirmation Bias
I've been meaning to link this excellent post by neo-neocon:
... the entire demographic “unmarried women,” ... seems absurd to me. It’s a term used by various research groups in studies, but it describes a conglomeration of women so disparate as to be virtually meaningless as a unit.
Think about it—”unmarried women” consists of women who have never been married (mostly younger ones, who will probably become married in the not-too-distant future), divorced women (many of whom will remarry, sometimes briefly and sometimes long-term), women with children and without, and widows (mostly older, most of whom will never remarry). What do these women have in common, besides being women, and besides being at least temporarily single?
As for unmarried women voting for Obama—whatever their reasons—some of this can be explained by the fact that they are predominantly young.
I also commend to you the linked study on why women file for divorce at the end of her post.
I do research for a living and there are two things I've learned over the years:
1. People love simplistic answers, especially if they happen to affirm what they already believe.
2. It's extremely hard to control for all - or even most - factors that influence outcomes. Many people dispense with this requirement by simply pretending they don't exist. That's not a strategy for success unless the purpose of your "study" is to confirm a pre-existing bias.
There's an old saying that goes something like, "Never attribute to malice what can be explained by stupidity". When surfing the Intertubes, I often think to myself that this old maxim ought to have a corollary:
Never attribute to a single, poorly defined factor what could just as well be attributed to numerous other factors you haven't controlled for (or even considered). That ought to go double for anyone who just found a handy study that "conclusively proves" them right.
I'm not holding my breath.
Posted by Cassandra at August 9, 2010 08:27 AM
TrackBack URL for this entry:
I've seen some data lately that says that the one group that voted overwhelmingly for Obama was single parents (men and women both). They went for Obama by about 75%. Take that group away, and McCain wins the election. I'll try to dig up the link.
Posted by: Cousin Dave at August 9, 2010 11:13 AM
I did a lot of research on this a few months ago and saw the same thing.
Singles and single parents are also statistically far more likely to be young, progressive, and/or black (all groups that overwhelmingly voted for Obama). Of course it's much more fun to lump all women into the same category and ignore the fact that most single men voted for Obama too (not to mention the fact that there wasn't exactly a huge margin of McCain-over-Obama voters among marrieds of either sex) :p
Posted by: Cassandra at August 9, 2010 11:41 AM
Careful how you lump widows into the "unmarried" category.
I vividly recall one elderly lady tartly informing me that she was NOT single, she was a widow.
To her, the words "Single" or "unmarried" meant "old maid" and were insults to a perfectly respectable and long-married woman whose husband had sadly predeceased her.
She EARNED her title Mrs. Smith and was proud of it. I never forgot her. Don't you forget her, either.
Posted by: Joe Doakes at August 9, 2010 11:49 AM
That's an excellent point, Joe. I found the whole "single women" thing mystifying.
It looks like a predetermined conclusion in search of confirmatory "evidence".
Posted by: Cassandra at August 9, 2010 11:53 AM
I like studies. Especially the ones the environmentalists do. You know, the ones that take up 5 tons of paper every year in cut down trees.
Posted by: Ymarsakar at August 9, 2010 04:11 PM
I'm not the single woman they talk about in these statistics...
Posted by: Miss Ladybug at August 9, 2010 10:24 PM
"I've been meaning to link this excellent post by neo-neocon"
I read it when she posted it; she's really good.
Posted by: camojack at August 10, 2010 01:23 AM