Since the release of a Pew Research Center report last spring on the “Changing Religious Landscape,” media outlets have suggested that the declines in Church affiliation indicate that the United States is becoming a nation that has given up on God. NPR claims that Americans—especially young Americans—have lost their faith. Now, a study reported in The Guardian tells readers that the loss of religion may actually be a blessing. Reporting on a study published in the peer-reviewed journal Current Biology, The Guardian claims that “Religious Children are Meaner Than Their Secular Counterparts.”
Claiming that religious belief is a negative influence on children’s altruism, The Guardian quotes Keith Porteus Wood of the UK National Secular Society, who lauds the study as a “welcome antidote to the presumption that religion is a prerequisite of morality … this goes some way to undoing the idea that religious ethics are innately superior to the secular outlook.”
Maybe not. The seven authors of the study of religious belief and children’s altruism, all university-based international scholars, set out to examine the influence of religion on the expression of altruism by designing a “resource allocation task” that involved sharing. Employing “the dictator game” which assessed children’s propensity to “be altruistic” to others in sharing their resources, the researchers drew a dubious sample that they called a “cross-cultural sample” of children (n=1,170, ages 5-12) from Chicago (USA), Toronto (Canada), Amman (Jordan), Izmir and Istanbul (Turkey), Cape Town (South Africa), and Guangzhou (China). Oversampling Muslims by almost 2 to 1 over Christians, the sample included 23.9 percent of the households identifying as Christian (n=280), 43 percent as Muslim (n=510) 27.6 percent as not religious (n=323) with tiny percentages of Jews, Buddhists, and Hindus. Claiming that their sample represented the “world,” (as their title “The Negative Association Between Religiousness and Children’s Altruism Across the World” purports) the researchers attempted to determine “how the religious rearing environment fundamentally shaped how their altruism was expressed.”
To measure altruism in the study, the “dictator game” was chosen—a game in which children were shown a set of 30 stickers and told to choose their 10 favorites. They were told that the stickers were “yours to keep.” They were then instructed that the experimenter did not have the time to play this game with all of the children in the school, so not everyone would receive stickers. Children were then informed that they could give some of their own stickers to other children who could not play the game by putting them in one envelope and they could put the stickers they wanted to keep in the other envelope. Altruism was calculated as the number of stickers shared out of 10.
According to the authors, the findings of the study revealed that coming from a religious family is “not associated with increased altruism in young children.” Religious children refused to share as many stickers as the non-religious children. They state: “Our findings robustly demonstrate that children from households identifying as either of the two major world religions (Christianity and Islam) were less altruistic than children from non-religious households. Moreover, the negative relation between religiousness and spirituality and altruism changes across age, with those children with longer experience of religion in the household exhibiting the greatest negative relations.”
Worse, the researchers suggest that religiosity affects children’s punitive tendencies when evaluating interpersonal harm. In other words, children from religious families are “more judgmental” and “meaner” than those from secular households. Overall, they claim that their findings
Cast light on the cultural input of religion on pro-social behavior and contradict the common sense and popular assumption that children from religious households are more altruistic and kind toward others. More generally, they call into question whether religion is vital for moral development, supporting the idea that the secularization of moral discourse will not reduce human kindness—in fact, it will do just the opposite.
Indeed, it appears that this may have been the real research agenda here. But, we, of course, cannot prove that. What we can suggest is that there are so many problems with the research design, methods, analysis and conclusions that no one should take this study seriously. First of all, selection/sampling bias emerges at once in this study because of the variation in age of the subjects (there are huge developmental differences between a 5-year-old child and a 12-year-old child). There is also measurement bias arising from the process of using the kinds of indicators the researchers used to measure “altruism” and “meanness” and “religiosity” in young children from different cultures and religions. Religiosity was assessed by asking parents questions like, “How often do you experience the divine in your everyday life?” The response to questions like this stood as a proxy for religiosity in the child. There are also problems with procedural bias—the study required different researchers in different countries to collect data posing major threats to reliability and validity. There is also reporting bias in the ideological language the researchers used to report their findings. All of these sources of bias should certainly call for caution in interpreting the results of this fatally flawed study.
William Briggs, an adjunct professor of statistics at Cornell, provides an overview of the flawed design and statistical analysis of this study. He first criticizes the indicator used to quantify “altruism,” and then ridicules the “moral sensitivity test” that children completed, which suggests a serious “abuse of regression on the pseudo-quantified answers … this model has no real predictive value.” Briggs concludes that “nearly everything is wrong with it, start to finish,” and is especially dismissive of the “wild, over-reaching theorizing about cause.” Suggesting that “altruism was not measured, but kids sticking stickers in envelopes was,” and he asks: “How much influence did the researcher have, especially with the younger kids? Did kids stick stickers because they wanted to prove to the whitecoat that they were compliant or because they wanted to be liked or because they wanted to share? Altruism forsooth!”
This is not to suggest that the researchers made up the data. It is clear that they conducted the experiment and analyzed the data they collected. The problem is the problem that haunts all social science researchers—garbage in, garbage out. The sad part is that a lot of time and effort was wasted on a study that is of little value. A Weekly Standard article published last month called “Making it All Up,” helps put this useless study into context by pointing out that ten years ago a Stanford researcher named John Ioannidis published a paper called “Why Most Published Research Findings are False.”
Indeed, an article published in the August issue of Science revealed that in an analysis of 100 studies from some of the most prestigious journals in social psychology found that nearly two-thirds of the experiments described “did not replicate.” This means that when scientists attempted to repeat these studies—using the same techniques, the same conditions, and analysis—they could not obtain the results described by the original researchers. The authors of the Science study concluded that, “This project provides accumulating evidence for many findings in psychological research and suggests that there is still more work to do to verify whether we know what we think we know.”
That is certainly an understatement. We need to begin to question these bogus studies that are done for ideological reasons. While there is nothing that can be learned from the study describing “mean religious children,” we can begin to look more critically at the ways in which we continue to believe the kind of junk science Current Biology is presenting.