A new international study of healing prayer published in the Southern Medical Journal suggests that prayer for another person’s healing can be effective if the one praying is physically near the person being prayed for:
Subjects exhibited improved hearing and vision that was statistically significant after [proximal intercessory prayer] was administered. Two subjects with impaired hearing reduced the threshold at which they could detect sound by 50 decibels. Three subjects had their tested vision improve from 20/400 or worse to 20/80 or better. These improvements are much larger than those typically found in suggestion and hypnosis studies.
Brown recounted that one subject, an elderly Mozambican woman named Maryam, initially reported that she could not see a person’s hand, with two upraised fingers, from a distance of one foot. A healing practitioner put her hand on Maryam’s eyes, hugged her and prayed for less than a minute; then the person held five fingers in front of Maryam, who was able to count them and even read the 20/125 line on a vision chart.
(Via: @Ariella09 on Twitter)
August 5th, 2010 | 2:25 pm
Normally, even though I’m not religious, I enjoy reading commentary at First Thoughts, because the bloggers are probing and interesting. This study is really falls short of the standards I normally see applied to writing here. Check out the methodology of this study, and you’ll see that it doesn’t hold water.
Here’s how they tested their hypothesis. After prayer was administered:
“Those who prayed then asked recipients whether they were healed. If recipients responded negatively or stated that the healing was partial, PIP was continued. If they answered in the affirmative, informal tests were conducted, such as asking recipients to repeat words or sounds (e.g. hand claps) intoned from behind or to count fingers from roughly 30 cm away. If recipients were unable or partially able to perform tasks, PIP was continued for as long as circumstances permitted.”
In other words, it was impossible for prayer to fail this ‘test.’ If subjects didn’t improve, prayer simply continued until the patients reported they felt better.
Especially when the details of the study are behind a paywall (as is the case here), it is irresponsible for First Thoughts to report the conclusions without reference to the invalid methodology. I hope you consider adding an update to this post to clarify.
–Leah @ Unequally Yoked
August 5th, 2010 | 2:53 pm
In other words, it was impossible for prayer to fail this ‘test.’ If subjects didn’t improve, prayer simply continued until the patients reported they felt better.
I haven’t read the study myself but it seems obvious how the test would fail: If there was no evidence of improvement in the subjects.
What is to account for the statistically significant results? I know PZ Meyers has tried to answer (“waking people up could generate an increase in blood flow to the head”) but it’s the usual sort of nonsense that keeps him from being taken seriously by his scientific peers.
There is certainly room for criticism of the methodology. Every study is flawed and I suspect this is one is probably even more flawed than most. But how can we explain the physical improvements? It’s not akin, as PZ claims, to “splashing water on your face” (honestly, how does anyone take him seriously?). We can say that more study is needed, but it seems that the results were clearly effective in this situation.
August 5th, 2010 | 3:06 pm
Thanks for the response, Joe.
You’re right to say that the study fails if there is no significant improvement in subjects, but the trouble is, the methodology eliminates the possibility that the results could be statistically significant. Improvement must be improvement relative to some baseline value. In this study, there are no controls that could establish such a baseline. Normally, scientific studies are very careful to include placebo treatments and untreated groups because there is plenty of evidence that the knowledge that one is receiving treatment is enough to cause mild improvement in many cases.
If I were to give sugar pills to the subjects instead of prayer and simply stayed in the room, continuing to give them placebo treatments until they reported they felt better, I could not claim to have evidence that sugar pills relieved their suffering. I’d need to have a separate group that did not receive pills but was also frequently asked if they felt better than previously. If the sugar pill subjects improved more than the subjects with no treatment, this would be an improvement I could ascribe to my treatment.
August 5th, 2010 | 3:11 pm
Even if there were controls, as I outlined above, the metric used for improvement was still too flawed to be useful.
Again, the methodology clearly states that treatment was applied until success was achieved. Imagine if we were testing if prayer could affect the outcome of a coin flip and the desired outcome was three heads in a row. I’m sure you would agree that it would not be a valid test if you and I simply sat down, and I flipped a coin while you prayed. I would write down the heads/tails sequence, and continue until I flipped three heads in a row, at which point you and I would agree that prayer worked.
Real scientific studies determine length of treatment and when data on improvement will be gathered before treatment begins, so we can’t preferentially sample only during relatively better moments (especially for diseases whose symptoms are variable).
August 5th, 2010 | 3:31 pm
Improvement must be improvement relative to some baseline value.
I agree. But wouldn’t the baseline ratio be no improvement? The researchers could have tested untreated groups, but wouldn’t the result be the same as “no improvement?”
As for the placebo treatment, I’m not sure how that could be tested in this situation. Would the placebo be someone claiming to pray for the patient, but not doing so?
If I were to give sugar pills to the subjects instead of prayer and simply stayed in the room, continuing to give them placebo treatments until they reported they felt better, I could not claim to have evidence that sugar pills relieved their suffering.
This seems reasonable. But would it require the same researchers who carried out the prayer study conduct the placebo treatment? In other words, why not just see what the results of such studies have been done in the past (surely there must be one) and compare it to the results.
From the press release, that sounds to be the approach they took. It says, “These improvements are much larger than those typically found in suggestion and hypnosis studies.”
Again, the methodology clearly states that treatment was applied until success was achieved.
I’m not sure why this is a flaw. If a pharmaceutical study said that the medication was taken until success was achived, we wouldn’t say it was flawed. It would mean that, with the proper dosage, the treatement was efficiaous.
Imagine if we were testing if prayer could affect the outcome of a coin flip and the desired outcome was three heads in a row. I’m sure you would agree that it would not be a valid test if you and I simply sat down, and I flipped a coin while you prayed. I would write down the heads/tails sequence, and continue until I flipped three heads in a row, at which point you and I would agree that prayer worked.
I’m not sure if the analogy holds. Since the tests were only recorded once, a more accurate analogy would be for me to pray as much as I wanted before the coin was flipped. But after it’s flipped, I can’t change the outcome.
Real scientific studies determine length of treatment and when data on improvement will be gathered before treatment begins, so we can’t preferentially sample only during relatively better moments (especially for diseases whose symptoms are variable).
Perhaps so. But the results seem to be clear that statistically significant improvement was detected. We can dismiss that because the methodology didn’t meet certain standards, but we’re still left with the question of what caused the improvement. I’m sure the people whose hearing and sight improved aren’t all that concerned that the methodology didn’t meet the standards of “real scientific studies.”
August 5th, 2010 | 3:32 pm
No matter who designs and conducts “prayer” a prayer experiment, I always wonder what possible good outcome could come from it. I wonder this primarily because I am under the impression that the primary purpose of prayer is not to procure some temporal end, but for the health of our souls. How on earth could anyone measure that?
Another bothersome element of such experiments is that in some way or another, God is being viewed in a manner that suggests that He is a celestial vending machine: as though we can evaluate whether this machine “works” like we hope or expect it to by imputting the “right” amount or type of change (prayer) into it. If the temporal outcome being prayed for (really, being demanded) doesn’t come to fruition, then it is claimed that prayer does not “work” (and, depending on who is interpreting the data, as further ‘proof’ that God does not exist).
Lastly, I have a hard time believing that God would allow Himself to be manipulated in such a way as to provide “proof” (in the form of temporal ends) in order to satisfy an experiment, regardless of the aims or methodology of the experiment or its designers. What in the Judeo-Christian heritage concerning prayer would lead anyone to believe He would?
Cordially,
GR
August 5th, 2010 | 4:10 pm
No matter who designs and conducts “prayer” a prayer experiment, I always wonder what possible good outcome could come from it. I wonder this primarily because I am under the impression that the primary purpose of prayer is not to procure some temporal end, but for the health of our souls. How on earth could anyone measure that?
I’m generally under the impression that primary purpose of praying for someone’s health is to procure health. This seems to me to line up pretty well with instances in the Bible where Jesus & the apostles prayed for healing for people.
If I want to improve the health of someone’s soul, I pray for their soul, not their ears or eyes.
August 5th, 2010 | 5:03 pm
Thanks for the reply. I’ll take a turn at the italics and answers style.
But wouldn’t the baseline ratio be no improvement? The researchers could have tested untreated groups, but wouldn’t the result be the same as “no improvement?”
As for the placebo treatment, I’m not sure how that could be tested in this situation. Would the placebo be someone claiming to pray for the patient, but not doing so?
Nope! Placebo effects vary by disease and type of treatment, but any treatment could presumably score pretty big gains (especially in the short term). To take only one example, pharmaceutical treatments for depression only edge out placebos by a small margin.
This prayer study had better design, since patients were not aware whether they were being prayed for. Sometimes the control treatments end up sounding pretty silly (possible controls: have atheist pray, have people pray to the Greek Pantheon), but they are important to guard against error.
. If a pharmaceutical study said that the medication was taken until success was achieved, we wouldn’t say it was flawed. It would mean that, with the proper dosage, the treatment was efficacious.
Nope, we would say the study was flawed. Studies that are determining appropriate dosage assign patients randomly to different levels of treatment and the effects are measured for each case. This is another reason I’m dubious about the claim of significance. The sample size was 24, which is already very low for a chance of getting significance (in fact, it is too low to be eligible for some kinds of statistical tests).
If the study design was dosage dependent, each dosage group would need to be analyzed separately, which would probably push the sample size too low for any kind of test.
[in re my coin flip analogy] I’m not sure if the analogy holds. Since the tests were only recorded once, a more accurate analogy would be for me to pray as much as I wanted before the coin was flipped. But after it’s flipped, I can’t change the outcome.
In the study, tests were not only recorded once. The researchers were sampling continuously during the prayer period (i.e. asking every few minutes if the subjects were healed). In my coin-flip analogy, you’d be sampling continuously from my flips, needing for one success to occur during prayer to declare success. As you said, this is not valid.
August 5th, 2010 | 6:39 pm
Tim:
I suppose I am not into “temporal results” kind of prayer – particularly the kind of prayer that has procuring a temporal result as its ONLY motivation, and especially the kind of impersonal and “let’s put God to the test”-type of prayer that “prayer experiments” invariably employ.
Given the earthly fates of Christ and the lion’s share of those apostles closest to Him – 10 out of 11 of the post-Resurrection disciples (only John was not tortured and executed), Paul, and James the Just – and taking into account the various Catholic meditations throughout Church history on the redemptive value of suffering, I cannot come to any other conclusion than that God’s first and foremost concern is for the perfecting of our souls, well above and beyond any concern for our temporal physical health or preservation. So, while I do pray for myself and others to be healed, I do so knowing that God may have something else in mind that would better serve His ultimate purpose for whomever I’m praying for: to live with Him forever. Therefore, I cannot and do not apply a “measure for success” of prayer based on whether the temporal result I petitioned for comes to fruition.
—————————-
Leah:
You wrote:
This prayer study (link) had better design, since patients were not aware whether they were being prayed for.
This was not the case for all 3 groups. Quoting the 4th paragraph of the linked article:
“STEP investigators enrolled 1,802 bypass surgery patients from six hospitals and randomly assigned each to one of three groups: 604 patients received intercessory prayer after being informed they may or may not receive prayers (Group 1); 597 patients did not receive prayer after being informed they may or may not receive prayer (Group 2); and 601 patients received intercessory prayer after being informed they would receive it (Group 3).”
Furthermore, one of the study’s authors noted a particular difficulty in maintaining the parameters of the test in groups 1 and 2:
“One caveat is that with so many individuals receiving prayer from friends and family, as well as personal prayer, it may be impossible to disentangle the effects of study prayer from background prayer.”
Moreover, the “Prayer Study” raises the very difficulty I mentioned at the end of my initial comment: I have a hard time believing that God would allow Himself to be manipulated in such a way as to provide “proof” (in the form of temporal ends) in order to “pass” an experiment, regardless of the aims or methodology of the experiment or its designers.
This is why I do not take the results of any “prayer experiment” as a data point either in support of or against the question of whether God hears and answers (in His way and in His time) prayers.
Both of you take care and God bless,
GR
August 5th, 2010 | 7:12 pm
more bad science. Placebo effect/ the power of suggestion is powerful, but not religion but an innate power of the mind.
and bad religion: As CSLewis wrote: Aslan is not a tame lion.
August 12th, 2010 | 12:53 am
Some thoughts.
GeronimoRumplestiltskin: “I have a hard time believing that God would allow Himself to be manipulated in such a way as to provide “proof” (in the form of temporal ends) in order to satisfy an experiment, regardless of the aims or methodology of the experiment or its designers. What in the Judeo-Christian heritage concerning prayer would lead anyone to believe He would?”
I’ll be frank: If this is the case, then this God is exceptionally capricious and cruel. Why? Because he is perfectly willing to (to some degree or another) improve your health if you pray or are prayed for, but refuses to if he knows anyone is keeping careful track of the situation. Presumably, people outside scientific experiments do benefit from prayer in a way exceeding random chance. So those who were enrolled in these tests just had bad luck and/or poor judgment. See what I’m saying? It doesn’t matter how horribly doubtful or selfishly inquisitive the scientists are — if God is “refusing to be tested”, then he’s punishing people of faith just for being part of an experiment.
Christians are free absolutely any time to stop saying that people can or should pray for others’ health, because God does not bother with such “temporal” things. But in the meantime, a God who supposedly does answer such prayers, in a real way, yet refuses to do so when anyone is “testing him”, is either capricious or nonexistent.
“Aslan is not a tame lion.”
In this specific case, it would appear that God is being the tamest of lions, because, since prayers to him aren’t outperforming the absence of prayers, he is behaving exactly as though he weren’t there. At the very least, a “wild lion” deity would, for example, make those who were prayed for do worse, just so we would all learn to fear his power. Better yet, of course, he would actually heal those who were prayed for more than others. (Better still, he wouldn’t afflict people with disease to begin with.)
Links
Blogs
Find Us
Contact