No Atheists in Foxholes, But What About Hospitals?

Debunking the study debunking the healing power of prayer:

Every day in the online journal Jewish World Review, publisher Binyamin Jolkovsky solicits prayers from readers for a friend who is ailing.

But if a recent study of the "medicinal power of prayer" is to be believed, then such pleadings aren't worth the trouble.

Researchers led by a Duke University Medical Center cardiologist divided a sampling of 748 heart disease sufferers, with about half receiving prayer and the other half not. Intercessory responsibilities, unknown to those who were ill, were divided among congregations of Christian, Jewish, Buddhist, and Muslim faith. The heart patients had received either catheterization or "percutaneous" treatment, and their conditions were tracked for six months. The findings were published in the Lancet, a medical journal.

Researchers discovered that those receiving prayer showed results that were little different from those not receiving prayer. Patients' conditions either improved or worsened in roughly equal degrees, regardless of divine entreaties or a lack thereof.

That's not surprising, because the study demonstrates a logical misunderstanding of God's nature, of prayer, and of His purposes.

The first obvious flaw is that believers from a variety of religions were asked to pray. Now this would be fine if the question was, "Who does God respond to?" Then the study might have determined that God reacts positively to one faith, while He disregards others.

But that's not why researchers divided prayer responsibilities. They exhibited no belief that an all-powerful, authoritative God would discriminate against certain religions. To the researchers, prayer to a Christian, a Jewish, a Buddhist, or a Muslim God were the same. They ignored each faith's claims about access to God, which are mutually exclusive of the others' (whether they admit it or not). To believe all paths equally lead to God, even if they contradict each other, denies that ultimate truth exists and that God is the source of it.

Another problem is that the researchers' premise assumed prayer is an earnest request that is always answered by an acquiescent God. If patients subjected to masked prayer didn't get better, then that demonstrated such activity was fruitless.

Assuming that God is obligated to respond to prayer in such a way diminishes His power and authority. It would show that He is at the mercy of man's demands, which would therefore not make Him God.

Nor can He be bound to respond affirmatively to all appeals, or else inevitable conflicts between requests would occur. For example, farmers in one area may pray for rain while vacationers might pray for sunny weather. Somebody's not going to get what they want.

For Christians, the real purpose of prayer is for believers to align themselves with God's will ("Thy will be done in earth, as it is in heaven" -- Matthew 6:10), rather than viewing God as a source to fulfill their wishes.

Also the prayer researchers, like many people, mistakenly believed that bodily health is the highest purpose for intercession. That approach presumes that the temporal condition is as good as life gets, and dismisses the hope of eternal glory.

If Jesus is your God, even the purpose for His healings in the Gospels was not the healings themselves. It was so that people would believe that He was the Savior and would lead them to eternal life. The healings were merely a means to that end -- a sign -- and an incidental benefit to His believers.

Indeed, the ultimate goal for God's uniquely created beings -- humans -- is to reach heaven. Whatever physical healing may occur is temporary in the ultimate sense. The purpose of prayer for the sick is to appeal to the One who truly has the power to save both the body and the soul.

Statistics are my life, so let's examine the study a bit, shall we?

If you're looking to test whether or not prayer heals, you would set the test up as follows:

Null Hypothesis: There is no difference between the healing rate of patients prayed for vs. patients not prayed for or patients prayed for heal at a slower rate than patients not prayed for.

Alternate Hypothesis : Patients prayed for heal at a faster rate than patients not prayed for.

This presumes we're testing the healing power of prayer---i.e., people who are prayed for heal faster.

Contrary to the author's claim, it is not unreasonable to simultaneously test another hypothesis---that the prayer of a certain religious denomination is more likely to heal than others. This is done to preclude having to test all over again to answer an obvious criticism---no religion believes all prayers are created equal.

This is done here with a stratified sample---basically, some rational basis is applied to the sampling plan subdividing the 748 patients into the various intercessory groupings. This is done all the time.

The real trouble with this approach, which I can't assess without seeing the study itself, is that 748 is a fairly low sample size when trying to describe such a huge population (potentially billions), particularly when you consider the disparities amongst strata. Even were it divided evenly, the highest minimum sample size would be 187. That should yield quite a large margin of error.

Moreover, how does one define the response variable here exactly---healing faster?

Were all these patients maintained in exactly the same conditions? Doubtful, over six months.

Did they have the same prospects? The same degree of illness?

Doubtful, of course.

The problem with these kinds of studies is that people differ in so many ways it is extremely difficult to find out what drives us. Look at all the people for whom aspirin is dangerous, for example, or for whom it is ineffective. We should be suspicious of any study purporting to demonstrate much of anything beyond the simplest causal relationships. I can't think of a much more complicated situation to assess than prayer.

When no statistically significant causal relationship is found between independent and dependent variables, one must interpret this in much the same was as an acquittal in a murder trial---it's quite possible that the suspect is guilty; we just can't prove it.

Many times, this is due to inherent flaws in methodology---it's pretty darned hard to construct a good study. The media, of course, is too ignorant to do anything but trumpet the press release of the special interest group which paid for the study.

As the author correctly notes, one must not presume to test God, nor presume that our will is God's will.