Today’s title was swiped, word-for-word, from a Live Science press release. This is important because the point I wish to make has to do with how the press and publicity treat papers; I have little to say about an actual paper.
And this is by necessity, since the paper being reported upon does not exist. Or, rather, it exists but will not be published until July, two months off, so far after the appearance of the press release that nobody in the media or public will remember to look the paper up when it finally shows. All that will be recalled were the old headlines.
Like the one appearing as our title today. Or the one, even, appearing at HotAir.com: “Confirmed: Atheists more motivated by compassion in charitable giving than believers are.”
Now pause and consider just these two headlines, the only words of a story likely to be read by most folks. What would your conclusion be? Why, that atheists are more generous, more compassionate than theists. This is striking because it goes against what many would have guessed to be true. Maybe there’s something to this atheism thing after all?
But suppose I told you I conducted an experiment with 100 atheists and 100 theists and asked them to self-report how much money they donated and the attitude they had whilst doing so. Say that 90 out of the 100 theists made charitable donations and that all claimed their activity was a duty; further say that just 10 out of 100 atheists gave but that each of these 10 reported doing so because they were moved by compassion. What would your conclusion be?
Obviously, “Atheists are more compassionate than theists.” Golly! This rendition is even true, in some political-spin kind of way. Not the way an ordinary citizen would understand it, of course, but ordinary citizens can’t be expected to comprehend difficult scientific concepts.
The first was to look at an old survey. Result: “compassionate attitudes were linked with how many generous behaviors a person was likely to report. But this link was strongest in people who were atheists or only slightly religious, compared with people who were more strongly religious.”
In other words “atheist” as defined by this first study meant “not-atheist” or “some atheist, some religious.” The lumping is suspicious and might—I say might—indicate their statistical correlation did not produce a publishable p-value when they compared actual atheists with actual theists. And just you examine what the study reports: the number of “generous behaviors” (whatever these are) was correlated with some definition of compassion. There is no word about the fraction of theists who engaged in “generous behaviors” versus the fraction of atheists who engaged in these.
In study number two:
101 adults were shown either a neutral video or an emotional video about children in poverty. They were then given 10 fake dollars and told they could give as much as they liked to a stranger. Those who were less religious gave more when they saw the emotional video first.
The description is not clear, but it appears that participants saw both videos but in random order. We’ll have to wait for the paper to see if just as many theists as atheists saw the “emotional video” first. Or whether they once again lumped some theists in with atheists to round out their numbers.
And did you notice that the fake dollars were given to a stranger? Why not real dollars to real children in actual poverty? Were these Monopoly money given to WEIRD college students? An odd, very unrealistic situation.
The third study:
[A] sample of more than 200 college students reported their current level of compassion and then played economic games in which they were given money to share or withhold from a stranger. Those who were the least religious but most momentarily compassionate shared the most.
Didn’t we see this or a similar game mentioned in yesterday’s post on the over-reliance of WEIRD people in academic research? Well, who knows. But how does one gauge somebody’s “momentary current level of compassion”? And what does fake money have to do with real-life giving?
The number of ways for this study to mean the opposite of the headline are many—but like I said, we’ll have to wait until we actually see the paper. There is at least good cause not to write the headline so boldly.
Live Science ends with a quote from Willer: “Overall, this research suggests that although less religious people tend to be less trusted in the U.S., when feeling compassionate, they may actually be more inclined to help their fellow citizens than more religious people.” Now why bring up that atheists are “less trusted”? What does that have to do with any of these studies? Might Willer be subject to a little confirmation bias?
Did you notice the qualifier? Atheists, when feeling compassionate, may be more inclined to “help” (with fake money, of course) their fellow citizens. What about when atheists are not feeling compassionate? Do they feel compassionate more or less often the theists? Who gives more? Who is in real-life more generous?
Aren’t these the real questions?