A publication called The WEEK published an article on the 2d which asked "Are Highly Religious People Less Compassionate?" - the issue has been covered in place like the Huffington Post. Each of the articles covered some "research" at UC Berkeley which conducted a series of experiments to discover compassion. One of the researchers commented "Overall, we find that for less religious people, the strength of their emotional connection to another person is critical to whether they will help that person or not," study co-author and University of California, Berkeley social psychologist Robb Willer said in a statement. "The more religious, on the other hand, may ground their generosity less in emotion, and more in other factors such as doctrine, a communal identity, or reputational concerns."
If you read the coverage of the research (I will admit that I have not read the article in the Journal Social Psychology and Personality Science to be published in the July edition) it has so many flaws one is limited only by the number of options on where to begin. One of the experiments had subjects see an emotional or neutral video and then received $10 fake dollars which they could give to a stranger. The non-religious people gave more after seeing the emotional video. The flaws in the methodology and assumptions on this study are huge. Did the religious people give less because they knew that simply giving money to a stranger, even one in apparent need, may not help the situation? (Is that really a doctrinal issue?) Did the subjects of the study perform differently because they were given "fake" dollars? The potential for absurd conclusions is gigantic.
One wonders why this research is important. There is plenty of evidence that both in terms of dollars contributed and in volunteer time that religious people contribute more to charitable causes. (See for example Why Good Things Happen to Good People). But the state of Publish or Perish in many universities encourages people to conduct this kind of silliness.