Wednesday, August 5, 2009

We Interrupt Our Regularly Scheduled Program for . . . Some Ramblings on Statistics

Let me tell you the most important thing I learned in my statistics class: correlation does not imply causation.

Today I was forwarded a warning about a certain common medical practice. Curious, I went to the website of the watch dog group included in the email. The site was slick and stylish, although a bit ominous with their rolling pictures of people who were victims of this common medical practice. I clicked on the link that promised to tell me why this common medical practice was wrong. The reason? The increase in the occurrence of this common medical practice over time also happens to line up the increase in occurrences of the unrelated X,Y, and Z diseases. So if you don't want diseases X,Y, and Z, you shouldn't participate in said common medical practice. By this logic, common medical practice also causes global warming. And an increase in the rise in the cost of fuel. And the decrease of Saturday morning cartoons.

I looked into it further because I was wondering why people link one particular disease with this common medical practice. From what I discovered, it had to do with an initial study in the 1990s. It was later discovered, however, that the primary investigator had been funded by lawyers working on a lawsuit claiming that this medical practice caused this particular disease, that he had been working on a patent for a rival variation of this procedure, and that the data now seems to have been manipulated and his collaborators outraged. Since then, there have been many studies that find no link between this particular disease and the common medical practice. But as the website I visited today shows, people persist in believing there is a link.

Oddly enough, I was just reading about a Christian denomination based in a movement led by a man who said the world was going to end in 1843. Despite the appearance that the world did not end in 1843, this particular denomination is now one of the largest Christian denominations. Another example of how belief persists even when seemingly disproven.

And that's ok with me. There are certainly lots of things that I believe in that lots of other people find irrational. Heck, I even find some of them irrational. On the other hand, many of them were irrational and unprovable to begin with which is why I don't worry about the tension between my two Faustian souls. That's why faith is what it is and doesn't need logical fallacies to support it. So to those against this common medical practice, I suggest abandoning the pseudo-science in favor of pseudo-religion.

In related news, this article almost persuadeth me to become a birther.

16 comments:

Schmath said...

First of all, this post and the article you linked to seem to be above my reading level. You are way too smart for me.

Second of all, how do you remember anything from statistics? Wasn't that like 10 years ago? All I remember is that the tests were in blue books and had study guides and that we thought we'd earned masters degrees in statistics by the time it was done and that there were lots of bassists in there and that Smee proved there were lots of potatoes in Idaho, but I don't remember any actual statistics.

Third of all, I'm curious what the medical procedure is.

And finally, does this mean I shouldn't worry about all those factors that seem to correlate with SIDS like having bumpers in the crib?

Vanessa Swenson said...

heh. nice. who sent you something related to the SDA?



I loved that article. It reminded me of the good old days in the JKHB.

linus said...

My favorite correlation is this one.

Save the pirates, save the world!

Vanessa Swenson said...

I've been meaning to ask you, do you particularly (not the BIG YOU) say "magister" with a Latin/German velar stop ([g]), or do you say it with a Vulgar Latin palatalized affricate? We'll just throw the soft French fricative out the door, I'm guessing.
Merci beaucoup.

Frau Magister said...

Schmath: First, you're the only person I've ever met who liked Kafka without reading the original so I'm sure it's not above your reading level. Did you get to the part about Obama being in league with the Norse god Loki?

Second, yes that class was almost ten years ago (so old!). But I remember the epic tests that took hours (literally, I think I spent at least 3 hours) and that there was a concentration of bassists. We must have learned a lot of statistics but you are right, the details are fuzzy, except that one of the guys who invented MANOVA was at BYU and we'd occasionally stalk his office to catch a glimpse of his genius. What I do remember was the father used the statistic that crime rates go up with ice cream sales as his example of correlation not implying causation.

Third, I'll give you a clue about the medical procedure (which I have tried to avoid naming because I don't want to offend anyone or get random hate comments): I discovered that TAMN of seriouslysoblessed.blogspot.com happened to blog about this very subject today! Correlation? Or is she reading my blog?

Finally, I'm certainly no expert in statistics or SIDS. The CDC appears to label the use of "pillow-like" bumpers, blankets, stuffed animals, etc, among the risk factors of SIDS. The theory behind that seems to be that SIDS could be caused by a lack of oxygen. But since the cause is still unknown and all we're left with are risk factors, which correlate with the known cases of SIDS, and results of preventative measures, like the Back-to-Sleep campaign which seems to correlate with a decrease in the number of SIDS cases. Some of the risk factors might be significant, some of them might be coincidental, and some may relate but not directly. Ultimately I would think one would have to decide what to worry about in consultation with one's spouse and the baby's health care provider.

Venom: I was actually reading about the Second Great Awakening on my own, so no one sent me anything about the SDA.
Viva le JKHB

Linus: Your Pastafarian sympathies are showing. But you know, Somalia does have the lowest carbon emissions and the highest number of pirates.

Frau Magister said...

Venom #2: I'm going to guess that I say it with the velar [g], but somewhat devoiced and palatalized. But a palatalized affricate sounds more fun.

Vanessa Swenson said...

thanks. i realized that i was saying a full-blown affricate. who knew that i was a latent 4th century Latin speaker?

Schmath said...

Holy cow! Longest comment ever! I can't believe you remember MANOVA. It sounds familiar. But I do remember stalking that guy's office. Rencher? Was that his name? Our teacher was obsessed with him. I think I used to have a higher reading level, but then I taught Junior High, so I only had to read easy stuff, and now I just read board books to Hanne, so I've retrogressed even further. By the way, is Hanne really a German spelling of Hannah? I can't believe I let Ryan talk me into that one. We might as well have named her Hannie. She's going to hate us one day.

The Old Cowboy said...

This is way beyond the old cowboy. There are a bunch of nutters around about many medical procedures. Remember the statistician who said that number of prenatal visits went up with bad outcomes, so therefore we didn't need prenatal care since it obviously caused bad outcomes. She was going to publish that one or try to anyway. Never mind that high risk pregnancies have more prenatal care because they are high risk.

REC said...

Oh boy you have us curious about "common medical practices". There are so many interesting medical "discoveries" that are rediscovered when another group offers money for research. "Columbus just discovered America again." is our response to them, along with, "Isabella still has lots of money." BTW speaking of "common medical practices" did you know your father has made his living taking out tonsils according to Obama. I have know idea who his statistician is but whoever he is he loves fuzzy math and typically isn't bothered by facts.

Troop 152 said...

"Aw, people can come up with statistics to prove anything. 14 percent of all people know that!"

-Homer Simpson

linus said...

So I went to seriouslyobsessed.blogspot.com , uh, I mean seriouslysoblessed.blogspot.com and I loved the comment "You should defiantly NOT vaccinate. You're kids might end up artistic if you do."

REC said...

Thanks for the link to Tamn's blog. (Even though I actually typed seiouslyobsessed the first time and Google took be right to it.) It is truly special blog. I felt I had wandered into paradise. I had this mad obsession to get a manicure and have my hair bleached. Your "Sarcasm Queen" title is in serious jeopardy.

Frau Magister said...

Schmath: Yes, a long comment, but you had a lot of questions. What was his name? Rencher? Rentschler? I can't remember and google and the BYU are no help. Alas. But Hanne is a German spelling of Hannah (although you might also see Hanna, Hanni, and even Hannah).

Old Cowboy: Good thing someone stopped her before she made a fool of herself in the medical community.

Mutti#1: The bean counters think pediatricians perform prostate exams. The politicians think pediatricians perform tonsillectomies. Like I've said, I'm all for taking health care out of the hands of the accounts, but I'm not sure putting in the hands of politicians is an improvement.

MMM, HansJuergen or TOK's Dad: Exactly. That's why you have to think critically about what people are telling you.

Linus: That was my favorite comment too!

Mutti #2: I would gladly give my sarcasm queen award to the genius who is behind Seriously, So Blessed.

Schmath said...

Ha ha. Imagine if we would have spelled it Hanni. Then the 2 % of people we meet who pronounce it right would have gone down to .0000001%.

By the way, this post has a lot of comments.

REC said...

I seriously does have lots of comments. What are the statistical odds for that?