Total Drek

Or, the thoughts of several frustrated intellectuals on Sociology, Gaming, Science, Politics, Science Fiction, Religion, and whatever the hell else strikes their fancy. There is absolutely no reason why you should read this blog. None. Seriously. Go hit your back button. It's up in the upper left-hand corner of your browser... it says "Back." Don't say we didn't warn you.

Friday, March 31, 2006

Heads and Tails

The Scene: Drek and his Sainted Girlfriend are listening to National Public Radio's "Morning Edition." The announcer has just reported that the only survivor of the Sago Mine Disaster has been released from the hospital. The announcer also reports that officials are at a loss to explain why he survived.

D'sSGF: It's because he's a cyborg.

Drek: Naw, don't you know? It's because god loved him. Which, of course, implies that god didn't love the other guys.

D'sSGF: No it doesn't.

Drek: Hmm?

D'sSGF: God loved the other men so much, he wanted them to come be with him.

Drek: Oh! So, god loved the man who lived less than the rest of them because he didn't die?

D'sSGF: No.

Drek: I'm confused.

D'sSGF: God loves them in different ways.

Drek: That's a heckuva difference.

D'sSGF: And god isn't going to change his plans for us.

Drek: So it doesn't matter if god loves you?

D'sSGF: Honey, it's early, can you please not annoy me until after breakfast?

Drek: Fair enough.

I don't recount this conversation just for humor value, or to annoy my Sainted Girlfriend (though that may well be a fringe "benefit") but rather because it illustrates a point. The funny thing about religious belief, about faith, is that people usually resist falsifying it. In the above example, my Sainted Girlfriend (who is vastly more positive about religion than I am) defended the notion that god loves human beings regardless of our apparent fate. So, somehow, both living and dying in a disaster like Sago could be taken as evidence of god's love. Now, certainly, I was the one to suggest the "god loved him" theory to explain the survival of Mr. Randal McCloy, but it isn't like I was the first one. Members of the Sago community seem to have latched onto the idea all on their own with McCloy's wife calling it a miracle and the road they live on having been informally renamed "Miracle Road." At the same time, I am also not the only one to note the uncomfortable implications of the "God loved him" account, as Mr. McCloy himself seems to realize. In reference to meeting the families of the other workers who did not survive:

“It’s a delicate situation, and it should be handled delicately. It’s not something you definitely want to dive right in,” he said. “I am going to choose to be careful about what I say and how I word things for the families’ sake. I just feel I should show them great respect.”

It's delicate because a bunch of talk now about how god loved Randal McCloy may not sit well with the families of the men who god, evidently, did not love quite so much. So, I can only commend Mr. McCloy for his good taste and self-awareness. Yet, I bring this up for a more fundamental reason: to once more express my reluctance to see science mixed up with religion.

Just as in the Sago disaster, where god's love can be used to explain both life and death with equal facility, scientific studies purporting to examine faith often fail for the simple reason that their results can always be explained away. In other words, results we approve of are accepted, while results we disapprove of are ignored. Falsification becomes impossible not because the test can't be constructed (although that may be the case) but because many people simply won't accept the results. Matters only become worse when we consider that studies mixing religion and science usually do justice to neither.

One supposed intersection of science and faith that has received a great deal of attention has been the role of prayer in healing. Many people believe that the prayers of others may, in fact, speed the healing process or improve a person's health. On the face of it, such an assertion would seem to be falsifiable- we put together randomized groups of people who are in need of healing, have other people pray for some them, keep everyone blind to the conditions they're in, and see what happens. Yet, even if such research were to be conducted, and conducted well, would people believe it? If scientists as a body said that praying for someone else does not improve their health, would people believe it? Probably not. On the other hand, if scientists said it did help people heal, would folks believe it? Probably so. We end up in a situation where, under the best of circumstances, only confirmation of pre-existing belief becomes feasible, and falsification is impossible. In such a case, it doesn't matter what the actual reality is- the accumulation of positive evidence through nothing more than random chance will provide (erroneous) support to an otherwise incorrect idea.

Yet, we are not operating under the best of circumstances, and studies of the role of prayer and health are, as a rule, poor. A reasonable scientific approach might be to start small- see if, maybe, prayer can alter bacterial growth in a petri dish. Then, work your way up to more complex phenomena. In reality, however, most prayer studies seem to start right at the top- looking at its effect on humans. Maybe that's becase god doesn't care about whether something lives or dies unless humans are involved, but that seems like an awfully strong assumption. And, of course, starting at this complex a level introduces a lot of confounding variables. In particular, there can be psychological confounds that may obscure the real relationship between prayer and health. For example, the pygmalion and placebo effects may replace the actual impact of prayer with the psychological value of believing people care about you. Despite this, however, we might still perform this research with adequate controls. Only, that largely isn't what is happening, and statistical artifacts are running rampant.

Take, for example, a study of the impact of intercessory prayer on recovery in a coronary care unit. It apparently found backward causation, meaning that the effect preceded the supposed cause. This is, to put it mildly, odd. More exactly, reanalysis of the paper data found that the prayer and non-prayer groups differed in terms of the likelihood that patients assigned to them would drop out of the study prior to the beginning of the prayer "treatment." Moreover, the difference in this tendency was significant at the p< .001 level, while the main effect of prayer (what the study was supposed to be examining) only just makes the p< .04 level. What does this mean? Well, either (A) the researchers biased the samples, which invalidates the results or, (B) the apparent effectiveness of prayer is a statistical artifact. Oops.

I'm prepared to believe that the above study was an accident, but not all cases of prayer/health studies can be so classed. In another case it's apparent that positive results were due to fraud on the part of one or more of the researchers- a fraud they apparently hoped to conceal with denials and a research paradigm of such inelegance that it would be impossible to deciper their mistakes. These efforts failed, and the subsequent scramble to deal with the fallout has been... interesting. Of course, research plows on and people continue to use these flawed studies to support the idea that prayer influences health.

Now, however, comes something more interesting. A recent study seems to imply that prayer does not work and that knowing you're being prayed for can even hurt. Now, I suspect that the effect they've identified is a statistical artifact, but here's the interesting question: How many of those who have used flawed studies showing an effect for prayer, will accept these results? How many will cheerfully give up their notions that prayer has a scientifically validated effect?

Probably none, and that's the problem. People believe in the power of prayer like they believe in god's love, and no amount of evidence will shake their faith. So, when we mix science with religion, we simply do violence to both.

Let science be science, let religion be religion, because otherwise it's just heads I win, tails you lose.

UPDATE: If you haven't seen it already, check out the letters to the editor of the New York Times regarding the prayer study. I think we can safely conclude, based on these letters, that I was totally right in my predictions. Drek- April 3, 2006.

FURTHER UPDATE: Sociology professor, blogger, and popular man with the ladies, Jeremy Freese has tackled this issue with his usual aplomb. I might have something to add in a later post, but for now, go check it out.


Anonymous Anonymous said...

based on these letters, that I was totally right in my predictions.

Wow, that's certainly a wonderfully arrogant example of confirmation bias. Of the 6 letters, only the 1 by Adjemian would find any and all results an example of God's love as you predicted. The others criticize the study's assumptions: prayer is not results-based (Hislop), different worlds/your conclusion (Xydis), the test was flawed (and a proper test is probably impossible) (Eckhoff), prayer is not about the patient but your relationship with God (Davis and Langwiser).

Tuesday, April 04, 2006 10:46:00 AM  
Blogger Drek said...

I applaud your efforts at turnabout, but I think you misinterpreted me, Anonymous. I said:

People believe in the power of prayer like they believe in god's love, and no amount of evidence will shake their faith. [Emphasis Added]

In other words, I was making an analogy here that disproving the power of prayer will involve the same problems as disproving god's love. I was not actually saying that all people would interpret any outcome as success, but rather that both things were approximately as difficult to falsify. An explanation can always be created to explain how a particular result either does support the notion that prayer helps or, alternatively, that a test was or was not valid. So, like god's love, prayer is more or less a fruitless area for scientific scrutiny. I think the letters support my overall point that we shouldn't be studying this scientifically because, fundamentally, people aren't going to accept the outcomes.

If that wasn't clear from what I wrote, I do apologize.

Still, it is nice to be referred to as arrogant again! It's been a while.

Tuesday, April 04, 2006 11:35:00 AM  

Post a Comment

<< Home

Site Meter