What Wilson did was to subject a group of over 100 undergrads (90% of them religious) to a stressful task - she videotaped them while they did a four-minute mock interview.
Before the interview, they each were given one of three different paragraphs to read. One was an encouraging prayer, one was an encouraging and motivational secular paragraph ("I can do this. Think of all the obstacles that I have overcome in my life..." etc). The third was the control group - they read about riding a bike.
The graph shows the results from systolic blood pressure. As you can see, both self talk and prayer were pretty effective. Although prayer just wins out, the difference isn't statistically meaningful. Similar results were seen from the questionnaire-based measure of stress.
So these results show that prayer is about as effective as self-motivation. For Wilson, this was a surprise. She was expecting prayer to be more effective, on the grounds that it creates the mental feeling of a wider circle of friends - a bigger and more powerful support network, if you will.
On the other hand, it does help to explain why people pray more in stressful environments. To be sure, motivational self-awareness might be just as effective, but prayer is what they know and intuitively turn to.
Belding, J., Howard, M., McGuire, A., Schwartz, A., & Wilson, J. (2009). Social Buffering by God: Prayer and Measures of Stress Journal of Religion and Health DOI: 10.1007/s10943-009-9256-8
This work by Tom Rees is licensed under a Creative Commons Attribution-Share Alike 2.0 UK: England & Wales License.