Stats Check, Revisited

Brent Kesler writes:
The statistic comes from the Sisterhood is Global Isntitute. If 6000 girls are genitally mutilated everyday, then in a 365-day year, that comes out to 2,190,000 per year, a lot more than 200,000. This has an error of nearly 1,000%. I must ask:

1) Is the Sisterhood is Global Institute in the habit of producing bad statistics? or
2) Do they merely pass on bad statistics without bothering to check facts? or
3) Is this a simple typo?

Laura Dellatorre responds:
Was it really necessary to list options one and two when the "error of nearly 1000%" almost certainly resulted from a typo? If you look to the bottom of the page, the source for the statistics is the United Nations Human Development Report. One has to doubt that the U.N. would provide statistics with such basic mathematical errors. And after all, typos do happen to the best of us: take Brent's misspelling of the Sister is Global Isntitute [sic]. Let's try to not let typos in web content sidetrack a serious debate.

No. I fully intend to sidetrack this serious debate--into another serious debate: the role of statistics in public discourse.

In response to her first question, yes. It was really necessary to list options one and two, even though it seems obvious it's a typo. Here's why I believe that, point by point.

1) Is the Sisterhood is Global Institute in the habit of producing bad statistics?

Personally, I believe that SIGI did not intend to misinform the public, and that the mistake was a simple typo. However, I will not assume that all statistics are objectively collected by disinterested persons. An unfortunate lesson I've had to learn in my life--there are people in this world who lie. Sometimes, they get together and plan their lies, to make them more believable. I do what I can to make sure I do not believe their lies.

For example, U.S. English is a PAC devoted to establishing English as the offical language of the United States. They routinely send out questionaires to people on their mailing list to gather stats to use in their lobbying efforts. The questions usually go something like this:

"Do you believe that immigrants should learn to speak English, the langauge 97% of Americans speak?"

"Do you want billions of YOUR tax dollars spent on bilingual ballots, tax forms, and license forms?"

The questions are designed to inflame the reader to answer the way U.S. English wants them to answer. Second, the questionaire is sent out to their mailing list, people who probably believe what U.S. English wants them to believe. Third, the respondents are self-selecting; they have to mail the questionaire back. The methodology makes no attempt to provide an accurate cross-section of public opinion. And yet, U.S. English uses these stats in their lobbying efforts, and to grow their mailing list, despite their screaming bias.

I don't like U.S. English or their policy goals. But a lot of other Americans think they're fighting the good fight. All must remeber, just because someone has good political goals, however you judge "good political goals," does not mean they'll refuse to cook the books. They'll ask misleading questions, and provide inaccurate information, all the better to support their agenda. This is done all over the political spectrum--no one group has the monopoly on irrationality.

2) Do they merely pass on bad statistics without bothering to check facts?

Again, I believe that SIGI did not mean to misinform the public. I also trust the U.N. statistic, though if their math is anything like their peace-keeping, I have cause for doubt. I believe it was a typo. So why do I whine?

First, the Sisterhood is Global resource page was last updated on November 25, 2001, as I write this. This means the mistake, whether the product of bad scholarship or just bad typing, has had nearly a year to spread into other brains. The counter on the SIGI homepage claims 25,772 hits, but neglects to mention when the counter started counting. Nonetheless, we know that this site gets significant traffic, and that for over a year, this significant traffic has read an inaccurate statisitic.

This wouldn't be a bad thing, if everyone read the statistic, thought "wait a minute--those numbers don't add up," tracked down the source, read the real statistic, and then passed on the accurate information. Ms. Dellatorre, however, did not notice, and passed on inaccurate information. Don't think I'm picking on Ms. Dellatorre--after all, I was already predisposed to distrust the statistic, yet I accepted it without question or pause, and did not notice the error until nearly a month later.

Dellatorre and I are reasonably intelligent people. As Dartmouth students, we are members of America's intellectual elite. Yet we missed this ridiculously obvious 1,000% error.

That's why I ask the second question. People hate to lose arguments, and they learn that whoever has more facts, tends to win arguments. So they collect statistics they read in the newspaper, see on TV, and stumble across on the Internet. They keep those statistics stored in their brain, just waiting for the next time they find themselves in an argument with some bigoted-or-misguided person-who-disagrees-with-me and open up the fire hose of reason on the flames of stubbornness. And as U.S. English demonstrates, there are people making false statistics to arm their unsuspecting comrades.

I must wonder, how many false statistics have been put out by organizations like U.S. English, and passed on as fact by well-meaning individuals who never checked the facts, or reviewed their methodology? I'm sure everyone in this forum has had a barage of stats thrown at them, and barely, or masterfully, deflected them with their own statistics.

I am going to make a bold assertion--many of the memes found awash in the airwaves and sleeping in black ink on white pages are mistruths, born either by innocents who don't know how to keep their thoughts tidy, or ideologues who care more for their struggle than for intellectual integrity. Much of the "common knowledge" and "convential wisdom" so many people appeal to is built upon errors and lies. If you don't believe me, then how to explain urban legends?

I do not wish to examine in full detail the sand in the foundations of some of our ivory towers, though I certainly welcome such discussion if others wish it. I do not intend, by drawing attention to Sisterhood is Global, to accuse feminists or feminist organizations of especially committing such errors. It was merely in their statisic that I found a mistake. I have since blitzed them to inform them that they may want to update their Resources page.

Instead, I intend to make a call for a sort of, with all due respect to Neal Stephenson, informational hygeine. We are not merely passive recipients of information. We are active carriers of that information. We choose to believe our various philosophys, and we can choose solid philosophys carved from the finest marble, or flimsy makeshifts that will blow away if ever the winds of reason touch it. So I implore both the scholars of this forum and the population at large: be careful where you put that statistic--you have no idea where it's been.

Post scritum

Junk Science Judo
Snowcrash

Posted by Brent at 7:16 PM |