Verified by 
I've spent the last eight years asking that question. Having satisfied myself—if not everyone I've met—that there's value in using empirical methods to study prayer, I've thought a lot about how to do it.
My research has taken me all over the United States, and to Canada, Brazil, and Mozambique. I've watched how people in different cultures pray for healing—who does it, what they say, what they do with their bodies, what they say they experience. And I've examined before-and-after medical records and written surveys; used medical equipment to test people's hearing and vision before and after prayer; and done follow-up interviews that span eight years.
Using an audiometer to test hearing before and after prayer. Copyright Candy Gunther Brown.
What have I learned? Prayer is the single most common form of complementary and alternative medicine. Reasons are not hard to find. When people feel sick, are in pain, or have received a discouraging medical prognosis, many people pray. Even people who are not otherwise religious reason that prayer might help, and it probably will not hurt. Public opinion polls tells us that a large majority of Americans—including medical doctors—believe that miraculous healing sometimes occurs. Scientific research has returned mixed results. Some studies conclude that prayer improves health, while other studies suggest that prayer may have no effect—or even lead to worsening health.
There are some who argue that scientists have no business studying prayer. What the late Stephen Jay Gould called the "nonoverlapping magisteria" of science and religion must be kept distinct. I believe that it is possible to respect the boundaries separating religion from science while asking empirical questions about prayer and health. And the fact is that people do pray for their health, whether or not their physicians approve. If prayer practices can affect health-for better or for worse-it seems to me that doctors, patients, and policymakers should all want to know. Once we have a clearer answer to the questions of whether and in what direction prayer affects health, we can look more closely at possible mechanisms.
Using visual acuity charts to test vision before and after prayer. Copyright Candy Gunther Brown.
How we go about designing studies of prayer matters. Most researchers have focused on distant intercessory prayer. I argue that it may be more useful to study what I call proximal intercessory prayer-in-person prayer for another person, involving emotional touch-the way many people actually pray for healing under natural (vs. laboratory) conditions. We need to pay careful attention to the theological beliefs and on-the-ground practices of people who pray as we design our studies. And, finally, we need to be creative and flexible enough to use a variety of methods, each of which is tailored to answer very precise questions about prayer and health. For instance, medical records can help us determine whether people claiming healing exhibited any observable improvement for which there is no obvious medical or natural explanation. Survey responses reveal how suffering people perceive the benefits of prayer. Clinical trials can show whether prayer practices result in measurable changes in health markers. And multi-year follow-up shows whether changes are temporary or enduring.
I am not suggesting that scientific testing can prove or disprove that "God" exists or answers prayer, or that the "healing power of prayer" is or is not real. But it is possible to study how people understand and practice prayer and to measure certain effects of prayer on health. This to me seems like exactly the kind of project that scientists should be eager to undertake.
