Monday, May 31, 2010

Good Math, Bad Math (science and technology)


REVIEWED BY: Ms. Cairo

MY RECOMMENDATION: YES, with reservations

AMAZON SUBSCRIPTION LINK: Good Math, Bad Math

WEB ADDRESS: http://scienceblogs.com/goodmath/

BLOG DESCRIPTION: Good Math, Bad Math is a blog which exists for two reasons: to ramble about the beauty of mathematics, and to track down the bozos who use bad math to lie, distort reality, and in general support bad arguments.

MY REVIEW: This blog isn't really about math per se, but about how people use, or misuse it. One of the problems with it is that the author doesn't post very often - for example the last post was May 3. In April there were six posts (most coming in a single five-day period at the end of the month). The blog is only 99 cents a month so if you've got the discretionary funds, why not.

Having said that, I don't know that I like the author. The sample post below is him mocking a student who dared to think that because his Christian college had had no murders in 4 years, it proved that Christian values saved lives.

And really, who is to deny it?

Critics mock us for our strict rules - like no dancing or drinking on campus, no members of the opposite sex permitted in your dorm room, nightly curfew hours - and the lack of a social atmosphere it creates. We have been the subject of books (God's Harvard), television shows, op-eds, and countless blogs who rant against our brand of overbearing right-wing Christianity that poisons society's freedom.

Yet, what is the cost of students being able to "express" themselves? Is that freedom worth the cost of drunk driving deaths, drug related violence, and love affairs turned fatal?


Now, I'm an atheist, but I believe in morals, and keeping the opposite sexes apart so there's no illegitimate births, and keeping alcohol off campus so a bunch of kids don't kill themselves - that makes perfect sense to me.

The author of this blog prefers to mock, and makes his case in a mathematical fashion, as you read below.

Good math and bad math must be exposed (for example the lies about man-made global warming, and the efficacy of wind machines should be exposed!) but picking on some kid because he's proud of his religious values? At least he's not advocating going around blowing up people of different religions...

On the other hand, I found his other posts very interesting and educational, so...

SAMPLE POST:
I know that I just posted a link to a stupid religious argument, but I was sent a link to another one, which I can't resist mocking.

As I've written about quite often, we humans really stink at understanding big numbers, and how things scale. This is an example of that. We've got a jerk who's about to graduate from a dinky christian college, who believes that there must be something special about the moral atmosphere at his college, because in his four years at the school, there hasn't been a single murder.

Yeah, seriously. He really believes that his school is special, because it's gone four whole years without a murder:

Considering that the USA Today calculated 857 college student deaths from 2000 to 2005, how does one school manage to escape unscathed? It's certainly not chance or luck. For Patrick Henry College, it's in our Christian culture.

Critics mock us for our strict rules - like no dancing or drinking on campus, no members of the opposite sex permitted in your dorm room, nightly curfew hours - and the lack of a social atmosphere it creates. We have been the subject of books (God's Harvard), television shows, op-eds, and countless blogs who rant against our brand of overbearing right-wing Christianity that poisons society's freedom.

Yet, what is the cost of students being able to "express" themselves? Is that freedom worth the cost of drunk driving deaths, drug related violence, and love affairs turned fatal?

There were 857 college student deaths in the five-year period from 2000 to 2005! Therefore, any college where there weren't any murders in that period must be something really special. That christian culture must be making a really big difference, right?

Well, no.

According to Google Answers, the US Census Department reports that there are 2363 four year colleges in the US. So, assuming the widest possible distribution of student deaths, there were 1506 colleges with no student deaths in a five-year period. Or, put another way, more than 60% of colleges in the US went that five-year period without any violent student deaths.

Or, let's try looking at it another way. According to the census, there are 15.9 million people currently enrolled in college. The school that, according to the author, is so remarkable for going without any murders in the last four years? It has 325 students. Not 325 per class - 325 total.

In other words, among a group making up less than 2/1000ths of one percent of the college population, there were no murders. Assuming that the distribution of violent deaths is perfectly uniform (which it obviously isn't; but let's just keep things simple), given that there were 857 violent deaths in the student population as a whole, how many violent deaths would you expect among the student body at his dinky christian college?

That would be a big, fat zero.

The fact that there were no violent deaths at his school isn't remarkable, not at all. But to a twit who's incapable of actually understanding what numbers mean, that's not the conclusion to be drawn. It's also not that the violent death among college students is actually remarkably rare. Nor is it that most college students will go through college without any violent deaths on campus. No - according to a twit, with 857 violent campus deaths over five years, the only reasonable conclusion is that there must be something special about the ridiculous religious rules at his college that prevented the great rampaging plague of violence from touching the students at his school.

I actually spent five years as an undergraduate at Rutgers University in NJ. During that time, there were no violent student deaths. (There was one death by alchohol poisoning; and there was one drunk driving accident that killed four students.) But zero violent deaths. Gosh, Rutgers must have been an absolutely amazingly moral university! And gosh, we had all of those horrible sinful things, like dancing, and co-ed dorms! How did we manage to go all that time with no violence?

It must have been the prayers of the very nice Rabbi at the Chabad house on campus. Yeah, that must be it! Couldn't just be random chance, right?

Ok, now let me stop being quite so pettily snide for a moment.

What's going on here is really simple. We hear a whole lot about violence on campus. And when you hear about eight-hundred and some-odd violent deaths on campus, it sounds like a lot. So, intuitively, it sure seems like there must be a whole lot of violence on campus, and it must be really common. So if you can go through your whole time in college without having any violence occur on campus, it seems like it must be unusual.

That's because, as usual, we really suck at understanding big numbers and scale. 800 sounds like a lot. The idea that there are nearly sixteen million college students is just not something that we understand on an intuitive level. The idea that nearly a thousand deaths could be a tiny drop in the bucket - that it really amounts to just one death per 100,000 students per year - it just doesn't make sense to us. A number like 800 is, just barely, intuitively meaningful to us. One million isn't. Fifteen million isn't. And a ratio with a number that we can't really grasp intuitively on the bottom? That's not going to be meaningful either.

Bozo-boy is making an extremely common mistake. He's just simply failing to comprehend how numbers scale; he's not understanding what big numbers really mean.


RECENT POSTS:
--Big Number Bogosity from a Christian College Kid
--The Danger When You Don't Know What You Don't Know
--Iterative Hockey Stick Analysis? Gimme a break!
--Finger Trees Done Right (I hope)
--Friday Random Ten, 4/23/2010

_______________________
Ms. Cairo writes several blogs including:
Seaborn: Oceanography Blog
Star Trek Report: Space Sciences
Topical Murder and Dated Death

No comments:

Post a Comment