Thursday, May 01 2008

Lies, Damn Lies, and More Damn Lies

As Wonderduck has pointed out, Aria episode 10 has been released and downloaded, and I almost can't wait to start watching.  But first, some serious business.

My mother and I do not see eye to eye on just about anything political.  She's staunchly anti-war, and unabashedly so, and we occasionally end up verbally sparring over the news.  There's no bitterness about it when it happens, except perhaps from my father, who holds his political opinions generally very close to his chest.

Today, she had some questions for me as a representative of the other side of the debate.  At the weekly historical society meeting, a member had been passing around copies of an e-mail purporting to list military deaths by year since 1980, including a helpful total of 14,000 for the Clinton administration, and my mother thought that that number looked fishy.   She's a history major, and tends to approach her chosen interest with the same degree of obsession I show towards my interests, and is good at spotting unusual bits that don't fit together.

The number looked fishy to me, as well, as I remember some pro-military bloggers citing a number somewhere in the 7,000 range for the same statistic.  I suggested that she check the information on line, as there was a handy web page listed.  She was worried that if the web page was a far-right site that she couldn't trust the information from.  So I went and looked for her.

The domain for the address given in the e-mail,, looked familiar, and when I went there I remembered why.  It's the page for the Federation of American Scientists, and I'd used it to do research on Warsaw Pact military equipment for a wargame.  They seemed largely non-partisan and trustworthy, and were a good reference source for research on international politics and military affairs.  I then checked out the specific document listed in the e-mail, the CRS (Congressional Research Service) report on American War and Military Operations Casualties: Lists and Statistics, published in mid 2007.

And, presto, my mother's question was answered on the first page, with a warning about an e-mail containing bogus statistics purported to be from the report.  The real statistics themselves follow, with 7500 US military fatalities between 1993 and 2000, which is roughly Clinton's time in office, so the ballpark figure I remember of somewhere in the 7000s fits.  I showed my mother the actual statistics and explained why I trusted FAS based on my experience with them.

My mother's now going to take the e-mail back to the group at the historical society next meeting and (knowing her) loudly announce the false information in the handout, and while I disagree with the opinions she holds, she will be in the right, because the statistics given were wrong.  Whether the statistics are true or false doesn't change the situation on the ground, but it does affect the credibility of the people providing the statistics.

If you see a statistic and it's too good to be true, check it out.  Check the primary source, if listed, and check statistics you know are trustworthy to compare.  If it checks out, you've got more ammunition in your arsenal of facts.  If it doesn't check out, you've just avoided shooting yourself in the foot.  Know that facts and statistics presented my be inaccurate or distorted.

If you're the sort of idiot on any side that makes up statistics to bolster your side, stop.  You're not helping anyone, you're only making everyone distrust anyone that believes differently and any facts that do happen to be true.  If you're doing it for that reason, you deserve to have a saguaro cactus shoved... [my ideal fate of those that deliberately screw with statistics is best not fully described]

Now, back to watching animated gondolas...

Posted by: Civilis at 08: 07 PM | Comments (2) | Add Comment
Post contains 758 words, total size 5 kb.

<< Page 1 of 1 >>
16kb generated in CPU 0.02, elapsed 0.0458 seconds.
39 queries taking 0.0304 seconds, 79 records returned.
Powered by Minx 1.1.6c-pink.