Mean religious kids?

atheism religion science

A few people have asked me what I think of the numerous news articles about a recent study purporting to show that religious children are less altruistic than non-religious children.

The Daily Beast sums up the findings with the headline “Study: Religious kids are jerks.” An article over at Groopspeak makes the audacious claim that “Major study shows that religious children are less moral than non-religious kids” Rhetoric like this is pretty difficult to justify in light of a closer look at the study (see below), but we live in a world where we automatically share the link to that story that says “Beer is good for you,” “Bacon isn’t so bad after all” or “Coffee wards off cancer.” We latch onto claims that the world is the way we would prefer it to be without so much as looking twice to see if it’s true.

The publication and publicity of the study, then, is bringing some predictable trolling out of the woodwork, revealing some rather old and tired misrepresentations of religion.

The way the study is being praised by some reveals the utterly ludicrous caricatures that some people have of religion. An article in the Guardian relates: “The report was “a welcome antidote to the presumption that religion is a prerequisite of morality”, said Keith Porteous Wood of the UK National Secular Society.” It’s a string of words I have heard before: That religion is not a “prerequisite for morality,” but it is a string of words only ever used by the anti-religious. There is of course an argument that the existence of God is the best way to make sense of moral realism, but that is nothing like this bizarre claim described by Wood. Religious people, at least for the main part, simply do not claim that being religious is a prerequisite for morality. I say “for the main part” because it’s always possible to find somebody saying something crazy. But they are hiding under a rock somewhere if they exist at all, and they certainly do not have a sufficient presence for there to be a widely held “presumption” that religion is a prerequisite for morality. The publication and publicity of the study, then, is bringing some predictable trolling out of the woodwork, revealing some rather old and tired misrepresentations of religion.

To his credit, Mr Wood did offer something reasonable as well: “We suspect that people of all faiths and none share similar ethical principles in their day to day lives, albeit may express them differently depending on their worldview.” No doubt. This is precisely what many religions teach, in fact. In the opening chapters of the book of Romans, St Paul is at pains to say that even people who do not have the Scripture nonetheless have God’s precepts written on their hearts. Christian moral theologians have never pretended that people who are not religious have no access to moral knowledge, and since God created the whole universe (and not just Christians), it should come as little surprise that we have much in common when it comes to moral belief.1

But what of this study? The authors themselves, as far as I can tell, make no bones about being non-religious people, and some of their previous work indicates a tendency to gravitate towards “debunking” various religious claims (although claims that are not overtly made by religion; for example debunking the phenomenon of seeing faces of Jesus in objects, explaining it as a natural phenomenon), as well as some rather interesting work in psychology and morality.

In short, this study is purported to have found that children from religious families are harsher, less empathetic and less generous than children from non-religious families. I have mixed thoughts on the study and the media noise about it (although there’s no detectable noise about it in my part of the world).

My first reaction is to roll my eyes at the predictability of the attention the findings are getting.

  • A study shows that adherents of organised religion have better mental health than other people (adding that those who were neither religious nor spiritual were the least likely to have a qualification beyond High School). The media doesn’t so much as murmur about it.
  • A study shows that less educated people are more likely to abandon religious faith than more educated people. Media silence.
  • A study shows that the further up the education ladder we go (proceeding from Bachelors to Masters to PhD), the greater the concentration of religious people we find. People who successfully make it past undergraduate study and into postgraduate study are more likely to be religious than those who don’t. No headlines.
  • A study shows that belief in God is actually the default natural state for human children, and that theism must be unlearned, rather than learned. Well, OK this one made a small amount of noise, but mostly in scientific literature. The general public doesn’t read that sort of thing, even, ironically, those who deride religion as being unscientific, while they, by contrast, are enlightened.

Many media outlets have selective hearing when it comes to scientific findings about religion.

But now a study shows that children from religious families are less likely to share stickers, and suddenly the story is passed around like a doobie at a libertarian meetup. It’s the old phenomenon of bias confirmation again, predictable and uninteresting. We naturally find our attention drawn to – and we’re only inclined to share – those tiny tidbits of information that reinforce our own prejudices. I mean sure, I did find the story somewhat interesting, but why didn’t the outlets that are sharing this story also find the studies mentioned above interesting?

So that’s my first thought – and my main thought. Many media outlets have selective hearing when it comes to scientific findings about religion, and that’s not particularly surprising. They ignore the good and trumpet the bad.

My second thought is “what’s the point here”? The people who seem happiest about the appearances of these stories are people who identify as atheists. They believe that theism is false. There’s no God and all religions are delusions of one kind or another. But how on earth do these findings, even if completely reliable (more on this below), offer any support for that contention? I can just imagine what people would be saying if the stories were saying that religious children were more generous: “Well, they’re only acting that way because of the myths they’ve been told about a God who will send them to hell if they don’t!” Maybe there are interesting stories to tell about why some children are more likely to give away their stickers than other children, but none of those stories are going to offer much insight on whether or not any religious beliefs are true. We can all probably come up with false views of the world that might somehow promote the giving away of stickers.

Relax. It’s downright interesting to observe patterns like these. Maybe there’s something to learn here (although as I suggest below, not as much as some might like us to think). If there are patterns of behaviour among children from religious homes that we’d rather not see, don’t you think it’s worth wondering why? The truth of your religion is not at stake, or even in question (not, at least, by anyone who is only concerned about what the evidence here is alleged to show). Don’t feel threatened.

Thirdly, although I’m a data and reporting analyst by trade (so I hopefully know something about making the move from numbers to conclusions), I’m no professional statistician. Don’t take this to be anything like a professional analysis. All the same, I have my suspicions about the breadth of the conclusions being drawn here.

For example, as a way of illustrating what the study’s authors mean when they describe children from religious families as harsher and more judgemental, the study tells us:

[C]hildren in Muslim households judged interpersonal harm as more mean than children from Christian (p < 0.005) and non-religious (p < 0.001) households, and children from Christian households judged interpersonal harm as more mean than children from non-religious households (p < 0.01). Moreover, children from religious households also differ in their ratings of deserved punishment for interpersonal harm (F(2, 847) = 5.80, p < 0.01, h2 = 0.014); this was qualified by significantly harsher ratings of punishment by children from Muslim households than children from non-religious households (p < 0.01). There were no significant differences between children from Christian households and non-religious households.

My summary: Children from religious homes were more indignant at interpersonal harm (Muslims more so than Christians), while children from non-religious homes were more chilled about it. However this only sometimes translated into children supporting harsher penalties for those who carried out the harm: Muslims supported much harsher penalties than Christians and non-religious, while there was no marked difference between children from non-religious homes and those from Christian homes when it comes to support for harsh penalties. This, of course, is also to simply ignore any protest from a Muslim person that their children were right to advocate harsher penalties for those who harm others. It should be clear that to leap from observations like these to claims about whether or not religion (in this case Islam) makes a person more or less “moral” is to leap right over the moral question altogether, never asking whether or not sterner penalties for offenders is right, just assuming that those who share your view are in fact morally superior. Even if you are right, I am more than a little suspicious that commentators on this study are oblivious to the way they normalise their own thinking and measure the study’s findings against it.

This raises the flag for me that the word “religious” is being used by many who comment on the article as though it covers a uniform phenomenon. Religious people are harsher, people will say (and are saying), even though the study found that there was no marked difference between the way children from non-religious homes and children from Christian homes advocated harsh penalties for those who harm others.

But still, religious people did more severely judge those who harm others. What the reader is supposed to take from these observations – or at least this is the impression I get – is that there is a disconnect here. Religious parents believe that their children are more concerned about the plight of others, but just look, they are mean and harsh, while the children of atheists are nicer, less judgemental.

I can understand this way of thinking – look at those condemning religious types, but I hope any reader will appreciate that just what counts as “sensitive to the plight of others” is going to be a bone of contention. Who are the “others”? What if the “others” are the victims of the meanness here? If being sensitive to the plight of others includes being more opposed to interpersonal harm (because interpersonal harm, surely, has victims whose plight we are sensitive to) then it isn’t clear why the above observation would warrant the children of atheists being assessed as more sensitive to the plight of others. If we are sensitive to the plight of victims, why would we not condemn interpersonal meanness? And if we are less inclined to condemn it, are we as sensitive to the plight of the victims of that meanness as those who do?

Finally, the test for altruism was performed by using the “dictator game,” which is described like this:

In this task, children were shown a set of 30 stickers and were told to choose their ten favorite. They were then told ‘‘these stickers are yours to keep.’’ Children were instructed that the experimenter did not have the time to play this game with all of the children in their school, so not everyone would be able to receive stickers.

The tendency to give away stickers here could, instead of being a manifestation of being more altruistic than other children, be a manifestation of a conformist disposition…

The more of their ten stickers a child gives away, the more altruistic he or she is deemed. Sure, it makes sense at a pretty simplistic level. Giving stuff away is altruistic. But – and I certainly have no expertise here – the dictator game is itself the subject of dispute among social scientists as a useful test of altruism. As Daniel Zizzo explains, a significant factor in determining what people will do in a dictator game is what he calls “experimenter demand effects.” Essentially, participants in the game are aware that they are part of the dictator game (or at least part of a game) and they realise that there are expectations of them, as well as, very broadly, what those expectations are. The tendency to give away stickers could, instead of being a manifestation of being more altruistic than other children, be a manifestation of a conformist disposition, doing what one hopes others expect or want one to do. Which is it? We will never know simply on the basis of what one does in the dictator game. One of the several reasons that dictator games remain popular in spite of their drawbacks, Zizzo remarks, is that “seemingly intriguing results can be obtained comparatively easily, and academic journals like to publish statistically significant results.” Here we have a case in point!


So what do I think of this study and the reactions to it? It’s not much of a blip on my radar. But when I do think about it, I think:

That it’s just one of many studies about the impact of religious belief on people, and the way that people choose to notice what research says is highly selective. When studies said many things about religion that we might think of as positive, the people who are reactive with apparent glee now were conspicuously quiet.

That people who are claiming some sort of victory against religion here are revealing an anti-realist approach to religion. Who really thinks that if Christianity (for example) is true, we should still believe that it’s false if it makes our children do X? Do we care that little for truth?

That the study is arguably careless when it comes to distinctions between different worldviews. “Religion” is at times treated as uniform (and this is certainly the case among people commenting on the study), while the authors perhaps fail to appreciate that “non-religious” may conceal a variety of different worldviews, and even if it did not, it would be a worldview along with religious views (rather than being a sort of neutral or “vanilla” worldview against which all religious worldviews combined can be compared). This has unfortunate and distorting effects on what is reported. For example when children from non-religious homes have the same approach to punishment that children from Christian homes have, while children from Muslim families advocate harsher punishment, we are told that children from “religious” families are harsher. Of course this is true if we squash together Christianity and Islam and call it “religion,” since that group overall will advocate harsher punishments than the non-religious. However, if we combine Muslims and the non-religious on the basis of the fact that they are non-Christian, we could say that non-Christians are harsher than Christians, since that group, overall, would advocate harsher punishments than the children from Christian homes would.

That the study’s writers seem to overlook fairly obvious competing interpretations of their data. If they are aware of these possible interpretations they do not address them in their published results. For example, the writers seem to find a tension in the fact that religious parents describe their children as concerned about the plight of others while those same children are judgemental of social harms. It is fairly easy to see, however, that being judgemental of social harms can be a manifestation of concern about the plight of others. Had they been sensitive to this possibility, some of the assessments about empathy could have been reversed when it comes to the differences between children from religious and non-religious homes. It is hard to see in a generous light why the researchers would have omitted this consideration, which makes me think they never considered it. I take this to be an indictment on their objectivity. On a related (but quite different) note, the study appears to suggest that its findings have something to say about a relationship between religion and morality. But at no point do the authors explicitly consider what, if anything, how the desire for harsh penalties for those who carry out social harms (as was exhibited by children from Muslim families) ought to be construed in terms of morality. In brief: What if the Muslim children are right? The authors, in other words, have just assumed that their own moral beliefs or inclinations are the obvious default position. They may be correct, but any inference from the attitude of these Muslim children to a claim about a relationship between Islam and moral behaviour is unscientific to say the least. Lastly the test used to determine altruism, namely the dictator game, may or may not have actually told researchers anything about altruism. They have evidently interpreted it as doing so, but as other experts in the social sciences have pointed out, the game makes participants vulnerable to the sort of influence that makes it difficult, if possible, to draw any conclusions about altruism rather than, say, how easily one is conformed to the expectations of authority figures (namely those who carried out the research).

I should add that, to their credit, there are good atheist minds out there who are not willing to trumpet the study’s conclusions just because it might work in their favour. Matthew Facciani is an example:

To summarize, this study was interesting, but a few fatal flaws severely limit the conclusions we can draw from it. These flaws are pretty obvious to anyone who has studied behavioral sciences (and may be why this paper was published in a biology journal and not psychology/sociology). However, I think this study does at least provide some evidence that atheist kids are not immoral monsters.

Relax, religious parents. Your kids aren’t ogres, or at least if they are there’s no scientific evidence that religion is to blame.

Glenn Peoples

Similar Posts:

If you liked this post, feel free to help support this project.

  1. “Moral theology” is that branch of theology that approaches ethics from an overtly Christian (or Jewish, or Muslim etc) perspective. []
{ 5 comments… add one }
  • David Hillary November 15, 2015, 3:45 pm

    Interesting study and I guess the saddest part is that somehow Christianity has lost its emphasis on the rejection of the punishment / deterrence / retributive justice / redemptive violence theory/theories. Clearly the Islamic culture and practice is much more into punishment / deterrence / retributive justice / redemptive violence than Christianity and secular culture and the sad thing is that Christianity still fails to provide as much contrast as it should.

    The Christian teaching on forgiveness, grace, love and peace is the great antithesis to punishment / deterrence / retributive justice and redemptive violence of the Old Covenant and of the ‘wisdom’ of the world. The wisdom of this world is that coercion, law and punishment is necessary for social order and to get people to refrain from wrongdoing and to pay their debts — that as long as the right cause exists and the right legal process is followed, and the legal limits are observed, one should judge one’s neighbour fairly and impose punishment or coercion to right wrongs and deter and control wrongdoing. The New Testament teaching is a real contrast to this whole line of thinking: judge not, fear not, eschew punishment, have and appeal to honour and communion, honour all men, forgive, love enemies, be like the Father, be perfectly good even in response to evil, do not repay evil for evil. The New Testament teaches that the fear and punishment idea indicates that one’s love is not made perfect. It teaches that the law kills and increases, not decreases, sin. It teaches that the law that promises freedom and justice through coercion is in reality a slavery system. It teaches that we have a ministry of reconciliation and peace. It teaches that if we want to reap righteousness we have to sow not just retribution/just punishment but peace. It teaches if we want peace with God we have to love and forgive our debtors — and yes that means letting them ‘get away’ with non-payment and with wrongdoing that they have not the will or means to compensate for.

    If Christians try to combine the punishment thesis with the love antithesis, they end up with a mixed message. The worldly wisdom recognises the merits of love and forgiveness in the right context and situation, but also insists that punishment and retribution and coercive law enforcement is also necessary in the background, and as, at the very least, a back up plan, or more likely, as a just default of rights and remedies. Sadly much Christian thinking and culture and practice is no different than the secular world and culture.

  • Glenn November 15, 2015, 4:46 pm

    “the saddest part is that somehow Christianity has lost its emphasis on the rejection of the punishment / deterrence / retributive justice / redemptive violence theory/theories.”

    I don’t even know that that’s part of the findings. After all, the children from Christian homes came out the same as the non-religious when it comes to punishing people – namely, at the least harsh end of the spectrum.

  • Jim Mason November 16, 2015, 1:46 am

    An excellent blog!

    However, I submit that the distinction between “religious” and “non-religious” is a false dichotomy. A Google search of the Internet for the definition of religion results in the following top three returns.

    From Wikipedia: “A religion is an organized collection of beliefs, cultural systems, and world views that relate humanity to an order of existence. Many religions have narratives, symbols, and sacred histories that aim to explain the meaning of life, the origin of life, or the Universe. From their beliefs about the cosmos and human nature, people may derive morality, ethics, religious laws or a preferred lifestyle.”

    From “1. a set of beliefs concerning the cause, nature, and purpose of the universe … 2. a specific fundamental set of beliefs and practices generally agreed upon by a number of persons …”

    From Merriam-Webster online: “an interest, a belief, or an activity that is very important to a person or group.”

    As can be seen, contrary to what is probably a widely held belief, these definitions do not include any reference to belief in a supernatural being. Although this can be part of a religion, it is not a required part. A well known example of a recognized religion that does not incorporate belief in a supernatural being is Buddhism, which “is a nontheistic religion or philosophy that encompasses a variety of traditions, beliefs and spiritual practices largely based on teachings attributed to Gautama Buddha, commonly known as the Buddha.” (Wikipedia)

    Since everyone has a “world view” or a “set of beliefs”, everyone is religious. The “non-religious” in this article are actually atheists, whose religion is, arguably, Humanism, which is, in effect, the worship of the human mind. So the partitioning in the study would be better described as “Humanist” and “non-Humanists”.

    The appreciation that everyone has a religion and, specifically, that Humanism is a religion, has huge significance with the respect to the subject of your doctoral thesis, at least for those who support the position “that the religious convictions that citizens in a liberal democracy hold should not serve as the basis for their support of public policy or their decision making in the public square”.

    (I would like to read your thesis. Is it available online?)

  • Matthew Flannagan November 21, 2015, 5:43 am

    So basically when you remove the spin, what the study shows is that that Muslims tend to hold less liberal or permissive views on punishment and religious people tend to view certain forms of wrongdoing more seriously than non religious people. Wow who knew?

  • Jason December 1, 2015, 6:09 pm

    If others who have analyzed the data are correct, the greater correlation is not between generosity (as defined by giving away stickers) and non-religiosity but between generosity and age. A twelve year old is more likely to give away their “favourite” sticker than a five year old. That might play into what you mention Glenn, that an older child is more likely to understand the expectations of the game, or it may simply be that a twelve year old doesn’t value stickers as much.

    In short sloppy research based on ill-defined concepts with conclusions that go well beyond what the data will support.

Leave a Comment

Remember: All comments should conform to the blog policy and you must use your real name. Comments that do not conform may be removed in whole or in part. You can review the blog policy here.

 Characters remaining