A Random World?

One of the most powerful components of the scientific method is to test hypotheses against a null hypothesis. Nearly everyone is capable of rational thought; but scientific thought is a discipline. Nearly everyone can reason from evidence to reach a conclusion; but in science, what we do is to contrast the evidence that we see for the hypothesis against the things that we would expect to see if the hypothesis is wrong. That is, against what we would expect to see at random.

And it is here that the scientific and the religious ways of thinking can perhaps most clearly be contrasted. For example, Herbert Benson and collaborators tested the hypothesis that God answers prayer. Actually, they tested the hypothesis that supplemental intercessory prayer would decrease the rate of relapsing back into heart disease.

In a scientific view of the world, you expect things to happen more or less randomly unless something is causing them to happen non-randomly. That is, absent some physical process, good and bad things will happen more or less equally. One of the things that this could mean in our everyday lives is that we should expect good and bad things to happen to us more or less randomly. As for the other humans with whom we interact every day, some are better and some are worse; and each person is a mixture of good and bad motivations. Therefore, in Benson’s study, one might expect patients suffering from heart disease to sometimes relapse and sometimes not. This does not necessarily mean that heart disease patients would relapse exactly 50 percent of the time; the actual rate will depend on the availability of good medical treatment as well as many other factors. But a scientific null hypothesis would state, in this case, that heart disease patients would relapse to the same extent whether they were being prayed for or not.

This is extremely different from the fundamentalist religious view. To a fundamentalist, the entire world is pervaded by evil, by the works of Satan, and that bad things will always happen all the time to everyone unless God specifically and miraculously prevents it. The fundamentalist null hypothesis is therefore 100 percent relapse.

Therefore, if you pray for someone to be healed from an injury or illness, and they recover, this constitutes evidence, or even proof, that God has intervened and blessed them miraculously, according to religious people.

Actually, there is no direct way to prove which null hypothesis—the approximately 50 percent relapse that scientists expect, or the 100 percent relapse that religious people expect—is correct. The only way to get around this problem is to have a control group. In Benson’s case, the experimental group of patients received intercessory prayer, and did not know it; and the control group of patients did not receive intercessory prayer, and did not know it. The percent relapse of the control group patients represents a measurement of the null hypothesis. It was in this manner that Benson and collaborators demonstrated that intercessory prayer had no measurable effect: one group had 52 percent relapse, the other had 51 percent, a statistically indistinguishable effect.

Fundamentalists have ignored this result and continue to insist that, unless you join their church and give them money, bad things will probably happen to you. God might allow them, or might prevent them. They insist that there is no such thing as God not doing anything; God either prevents bad things, or else bad things happen.

This is perhaps the most basic difference between a scientific and a fundamentalist view of the world: the scientific view that things happen at random unless they are caused, and the fundamentalist view that only bad things happen unless God prevents them. To a scientist, the world has a random background; to a fundamentalist, the world has a background permeated with evil.

I also posted this on my science blog.

The Search for Evidences, part two

One of the most famous recent pieces of scientific research was an article that appeared in American Heart Journal in 2006. The lead author, Herbert Benson, is a physician at the Mind/Body Institute in Chestnut Hill, Massachusetts. His co-authors included other doctors, Ph.D.s in medical sciences, nurses, public health experts, and theologians. This extensive team conducted what is perhaps the best study to date of whether or not prayer works.

In this study, Benson and co-workers identified about 1,800 people with heart problems, and who were scheduled to have a heart bypass operation, in hospitals across the United States. The investigators wanted to know if intercessory prayer would help the patients recover from surgery. The experimental design would seem to be very simple: some of the patients would receive supplementary intercessory prayer, and some would not. (You cannot prevent people from praying, but the investigators arranged for Catholic and Protestant religious groups to pray for the list of patients, in addition to whatever prayer from friends and family they were already receiving.)

But there is a confounding factor in this approach. If the patients who received supplementary prayer knew they were receiving it, this might have a psychosomatic effect upon them. They might feel better, and even recover better, simply in the knowledge that they were being prayed for. The prayer itself might have no spiritual or miraculous effect. Benson and colleagues were aware of this, and they designed their experiment to avoid this problem. They divided the patients into three groups, not two. Patients in two of the groups were told that they might or might not receive extra prayer. The patients in these two groups therefore could not know whether they were receiving supplementary prayer. Of these two groups, one received supplementary prayer, and one did not. The patients in the third group were told that they would receive supplementary prayer, and they did.

The two groups that were uncertain of receiving intercessory prayer had virtually identical rates of complication following surgery. This demonstrated that prayer, itself, did not cause healing in these patients. The group that received prayer and knew it had, as predicted, a psychosomatic effect, only it was the opposite effect from what was expected. These patients actually had more complications after surgery. This surprising result could not be explained, except that perhaps it was performance anxiety: the patients knew that they should recover from surgery, because if they did not, they were letting God down.

These results have left religious people, who believe that God answers prayer, scrambling. Their most common response was that God simply refused to go along with the experiment. God refused to heal the heart patients because He did not want anybody investigating Him. (This seems a strange reason for a God of Love to let somebody die.) Of course, nobody can say. The experiment tells us nothing about what God, if any, might have been thinking. All it shows is that prayer does not reliably work.

So, just as with the near-death experiences, we are left with a jarring lack of evidence that there is a God or an afterlife.