« Join Our Adventure | Main | Gellman Gets It Right »

June 29, 2007

Comments

This brought to mind the report last year that actually said, "Eating Chocolate May Halve Your Risk of Dying."

Would that it were as innocent.

(See http://www.newscientist.com/article.ns?id=dn8780&feedId=online-news_rss20)

Steve says something great here: "The lesson? Don't be impressed by statistics unless the conclusions drawn from them are true."

I was taught by one of my proffesors in nursing school who did her doctoral thesis on statistics that: "Always remember that figures can lie and liars can figure." (not that I'm calling anyone a liar)

She went on to give us a good example:

Statistically, there is a link between the hot dog sales at a famous baseball stadium and the water levels of the Mississippi River. The hot dog sales were highest when the river was measured at it's lowest. Why? Because the jump in hot dog sales had to do with the time of year, as do the water levels, however, one thing had absolutely NOTHING relevant to do with the other. Lori Vance, Lpn.

Exactly, Lori. Correlation does not necessarily imply causation.

You know, I constantly read feminist and center-left blogs that claim that as soon as abortion is illegal you all are going after Griswold and contraception. Regardless of the merits of the claim, it it not possible that giving folks another reason to practice contraception would lead to less abortions? What gives here? If folks really want a child they will seek to get pregnant and take their chances. If not, what's the harm in encouraging behavior that would arguably reduce abortions? What ox of yours is being gored here?

Alan,
I have several problems with contraception, only one of which has to do with my religous beliefs. One big issue for me is the behavior it seems to encourage. In my county, a little girl of 12 is allowed to walk into the health department and get birth control and even condoms without parental consent. She is not likely to be a good enough medical historian, nor is she mature enough to deal with getting help if she begins to suffer any ill side effects. She thinks and lives only in the moment, because of her lack of maturity. She thinks with BC pills on board, she can experiment to her little heart's delight. Try this one, a sexually abusive father getting his child on this medicine to reduce the risk of her getting pregnant from his abuse. But even if adults are using it, I still think it should not be used. It discourages self control, and encourages selfishness. It can come between a man and his wife in the area of intimacy. It also carries with it medical risks to the woman that clearly should give anyone who cares about women pause. Not to mention that it does cause abortion, it does not prevent fertilization. If a woman does not want to get pregnant, then she should NOT be having sex.

In a statistics class one learns that "There are three kinds of figures: lies, damned lies, and statistics."

I remember more but this is the best one liner and actually it came from a dean, not the professor.

People can have all the numbers and the data but still not be able to make sense of things. Statistics don't so much tell truth as they just merely summarize data.

Moving on from there, marketing and decision making are still two other fields.

There isn't an ox, Alan. I don't think in terms of oxen or teams or pet projects. I care about the ideas and whether people are thinking and communicating clearly. The point of the blog was simply to criticize one example of poor thinking and poor communication. The example could have come just as easily from a pro-life blog which concluded that "oral contraceptives never prevent ovulation" or the "fetus always feels pain in an abortion" or "the fetus has brainwaves at 6 weeks." We should be on the lookout for overstatement and neglect of data wherever these errors rear their ugly heads.

Hello all,

It is interesting to see stats used to push a point. I highly recommend reading A Mathematician Reads The Newspaper by John Allen Paulos. He does great job explaining the misleading use of stats in everything from the accurracy of pregnancy tests to "life promoting benifits" of x y or z.
D.

Paulos is excellent. "Innumeracy" is another worthwhile read by him.

Excellent point about considering other aspects of the risk equation.
Take for example the Gail Model developed at the National Cancer Institute that assesses a womans risk of getting breast cancer over five years and over a lifetime. It can be found here:
http://www.cancer.gov/bcrisktool/
You will note that question #4 asks about a womans age at the time of her first live birth.
That is because women who carry a pregnancy to completion earlier in life seem to have a lower risk of getting breast cancer during their lifetime and this statistical FACT is reflected in this equation.
Nulliparous women (women who have never given birth) have a higher risk of breast cancer.

Conversely women who terminate (abort) their pregnancy seem to increase their risk. The Gail model has not been updated yet to reflect this fact.

Hi Steve, fair enough, and if there is no connection that should be pointed out, however this sentence seems to indicate a certain animus towards contraception. Most folks use contraception to plan their families rather then eliminate all possibility of pregnancy. For that matter, if someone is insightful enough to know they would be a lousy parent and don't really want children then isn't it better for them to avoid the possibility?

"You have 1/5 the risk of death of carrying a pregnancy to term just for taking the pill? That's more than I expected. And what do you have to show for it? Nothing. Oh, I forgot - that's the point."

Hi Lori, we will just have to disagree on this. Family planning come with a middle class society.

what county do you live in? I hear about the twelve year old from time to time. I don't happen to believe that the availability of contraception encourages much adolescent activity as opposed to making it safer.

Well, if sexual promiscuousness is safer, then wouldn't be easier to justify? If an action is easier to justify, then won't more people end up engaging in that action? Many high-schoolers have had sex, and in college it simply becomes ridiculous. I don't reckon this would be the case if BC and abortion were at least less accessible.

Alan,
Thank you for your kind response. First, can I ask a question? Are you saying that we need "Family Planning" services to maintain a middle class? I'm just trying to understand your point. Could you clarify for me, please? In answer to your question, we live in a small county in a southern state, that I see in large part, to be very conservative. Which is why when I asked the local Health Department how old you have to be to get BC and condoms, I was shocked when she said "twelve". I have a twelve year old! They can get plan B as well. This is quite disturbing, as I picture one of my child's friends bleeding to death from plan B before anyone knows about it. Also, I know these local kids have said that ( in regards to sexual activity) "...it's all good, 'cause you can just go the Health Department to get protection and they won't tell on you.". Furthermore, kids made similar comments in my older child's health class at the high school. So, since I have seen first hand the results of availability of "protection", I can't deny them. Finally, It's never wise on the part of the adults to encourage kids to do something we know to be harmful by saying to them..."...since you will do it anyway, here is how you can be possibly free of the consequences.". This is never a good lesson to teach them. Lori V.

I recall when I was in Jr. high and high school several incidents among them a girl going to TJ for an abortion, another girl who was there and then she wasn't and the letterman who suddenly dropped out and was soon married and a father.

An STD or a pregnancy is a stiff price to pay, too stiff in my opinion, for an unwise decision and/or poor parenting. Evolution has given us a hormonal system that is predicated on our lives being relatively short and little education needed. We live in a world where maturity is earlier, life is longer and education well into our twenties is not unusual. I'm not impressed by notions that kids who think they are immortal will will back off from risky behavior because of potential consequences; some will, too many won't. I would rather they get old enough to reflect and shake their heads.

Yes Lori, I don't think it is possible to keep half your potential workforce pregnant or raising young children and maintain a middle class society. One of the first choices women make with increasing education and wealth is to limit family size. Large families are easier on a farm, not so much in an urban setting. Given their druthers most folks seem to want one or two kids and, based on my observation, most of us are barely qualified to raise that many. The ability to successfully raise and enculturate more than that number is a rare quality - the problems of the baby boom generation should have convinced us of that.

Hi Alan
do you have any evidence to support your claim about what evolution has given us regarding our hormones? Are you suggesting that our behavior is fatalistically determined by our hormones with respect to sexuality? Do defective hormones account for pedophilia? If so, why do we get so angry at pedophiles? Or is our anger likewise determined by hormones?

Hi Patrick, I think most folks got my point. You are really reaching here.

Alan
if "most folks get your point", then they too have abandoned reason on that point.

Hi Patrick, I think you may have a category confusion. Hormones deal with basic drives. The direction those drives take can be a bit more complicated.

Anger can in fact be driven by hormones and other drugs. A recent example is Chris Benoit.

I thought it was generally accepted that adolescents were hit suddenly with a flood of hormones, and dealing with that was part of growing up. You disagree?

The comments to this entry are closed.