William M. Briggs

Statistician to the Stars!

Here’s Why That Study Claiming Religious Kids Are Less Altruistic Stinks. Updates

Fig 2 from the article; colors inverted.

Fig 2 from the article; colors inverted.

Welcome Guardian readers. You should understand this poor study is one of thousands, all of which make similar mistakes. See the comment at the bottom of this article for more resources.

Heard about that scientific study which scientifically shows non-religious kids are scientifically more altruistic than unscientific religious kids? The Guardian summarized it thusly: “Religious children are meaner than their secular counterparts, study finds: Religious belief appears to have negative influence on children’s altruism and judgments of others’ actions even as parents see them as ‘more empathetic’.”

Scientifically speaking, this is crap. Here’s why.

The scientific science “study” The Guardian cites is the peer-reviewed article “The Negative Association between Religiousness and Children’s Altruism across the World” in the journal Current Biology by Jean Decety and a bunch of others. Biology? Never mind.

Authors gathered kids, 5 to 12, from the US, Canada, China, Jordan, Turkey and South Africa and asked them a bunch of scientific questions, scientifically quantified those questions, produced scientific statistics, and then made scientific propositions about the whole of the human race. Say, are there differences in behavior between 5- and 12-year-olds? That doesn’t sound like a scientific question, so never mind.

Here’s how to you can replicate their study at home. First, define altruism. Go on, I’ll wait.

Have a definition in mind? I’m sure it’s correct and matches everybody else’s definition in precise detail, details like no-greater-love, supreme sacrifice, kindness, patience, love, and so on, right? Well, maybe not, but never mind. Instead, think about how you would quantify your definition. Quantification makes your definition scientific. Science means unquestionable truth.

Was your answer about quantification the “Dictator game”? Like this (from the Supplementary description)?:

[C]hildren were shown a set of 30 stickers and told to choose their 10 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 the school, so not everyone would be able to receive stickers. Children were finally shown a set of envelopes and informed that they could give some of their stickers to another child who would not be able to play this game by putting them in one envelope and they could put the stickers they wanted to keep in the other envelope. Experimenters turned around during the child’s choice and children were instructed to inform the experimenter when they were finished. Altruism was calculated as the number of stickers shared out of 10.

Yes, this scientifically captures every possible nuance of the scientific concept of altruism, doesn’t it? Science science science science. Science. It must be science! Scientists wrote this, peer scientists reviewed it, and scientists nod sagely when reading it.

Now define “religiosity” for kids. I’ll wait again.

Have it? Ha ha! That was a trick question. The authors never assessed the “religiosity” of kids; they did it for the kids’ “caregivers” instead. How? The authors asked parents to name their religion. They also asked parents questions like “How often do you experience the ‘divine’ in your everyday life?” They took pseudo-quantified answers from these and combined them scientifically with a quantification of religious attendance and derived a complete scientific quantification of “religiosity.” This was assigned to each kid in the study.

After that, “Children completed a moral sensitivity task programmed in E-prime 2.0 and presented on ASUS T101MT Touchscreen computers…” My goodness! How scientific! An ASUS T101MT! Just think how dramatically the results might have changed had they used an ASUS ROG G752! Or an ACER C910-C37P!

You know what happened next. Wee p-values through the terrible abuse of regression on the pseudo-quantified answers. A picture showing one of these is at the top. Notice the wee p-values? That makes the findings scientific.

All those dots are the answers to the pseudo-quantifications for each kid. The flat surface is the regression (expressing this and nothing else: the change in the central parameter of a normal distribution representing uncertainty in “altruism”; did you think it was something different?). Notice almost none of the dots are near this flat surface? That means this model has no real predictive value.

Which, scientifically speaking, means this study is crap. And where I use “science” in the old-fashioned, pre-government-grant way.

Finally, no paper would be complete without wild, over-reaching theorizing about cause. The authors say their findings “contradict the commonsense and popular assumption that children from religious households are more altruistic and kind towards others”. Idiots everywhere are taking this literally. I’m too tired to make a joke about science. You do it.

This study is so bad that it’s good. I mean, it stinks to high heaven; nearly everything is wrong with it, start to finish. Yet it’s good because it takes so much effort to dissect, and the effort reduces the critic to such a sputtering mess that the criticism is bound to sound like an old fart yelling at the kids to get off the lawn.

Why I am Right Update P-values are rotten evidence for anything (click here to learn), (2) Regression is deeply flawed and not what you think (click here, here, or here to learn), (3) Probability models do not prove cause (click here, (4) Asinine studies like this are common (click here) or here). And don’t forget that altruism was not measured, but that kids sticking stickers in envelopes was. How much influence did the researchers have, especially with the younger kids? I mean, did kids stick stickers because they wanted to prove to the whitecoat they were compliant or because they wanted to be liked or because they wanted to share? Altruism forsooth!

And most importantly, don’t throw a temper tantrum and (for you men) cry like a little girl or say that I am a bad person. Show—exactly—where I am wrong and the authors right. If you cannot do this, keep your mouth shut. (I am too tired and too surly this morning to put this more politely).


Thanks to the many readers who pointed me to this study. For more resources about why this paper stinks, see my comment to Doubting Thomas below, or click Classic Posts above.


  1. Among the things that are truly fascinating about this study is that the researchers seem to believe that the children understood not only the “rules” but also the complicated set-up of the sticker game right away (and what of questions the kids would likely ask: Why aren’t there enough stickers to go around? Why am I playing this game? Aren’t you going to my sister’s class?).

    Also, the children may have been confused if the exercise was introduced as a “game” and schoolchildren are accustomed to games that rely on wit, skill, or chance. Choosing favorite stickers really doesn’t have a “game” quality about it.

    What is all that business about the researchers being “turned around”? Don’t they know that children are usually up to no good when their minder’s eyes or closed or heads or turned? Was there a way to stop inter-group sticker swapping? Also, a 12-year-old is almost 13, which is practically 30, and stickers may not necessarily appeal to students of this age.

  2. Doubting Thomas

    November 9, 2015 at 9:06 am

    Your review is certainly flawed. Some years will pass before we learn if this study is flawed. Its authors certainly are doing science, and they and some of their peers will continue to do so as they pursue this significant story.
    Those wee p-values you so inappropriately criticize, by the way, are in the knock your socks off category. In a former life, I was blessed when editors for a world class journal accepted some of my work with not quite so robust results.

  3. We also aren’t told how many how many out of the 10 stickers each child was willing to donate to a less fortunate child who got none. The Bible says to tithe 10%, so if each child was willing to donate just one, then that satisfies the religiosity test. The problem here is that left wingers expect you to give more than that. I’m willing to bet that if a child weren’t willing to donate at least half his stickers to another child, and as many as 8 or 9 if they were considerably more well off than the rest of the class, that qualified them as being mean.

  4. Briggs

    November 9, 2015 at 9:23 am

    Doubting Thomas,

    I submit (and I mean this in the nicest possible way) (1) You have no idea what a p-value is (click here to learn), (2) that your knowledge of regression is deeply flawed (click here, here, or here to learn), (3) that your knowledge of the purpose of statistics fails to realize that probability models do not prove cause (click here to learn), (4) that your saying my review is flawed without providing any proof of this is useless, (5) that my review showing where the study itself is badly flawed is correct and, if anything, generous. Further, saying this work deserves respect because it appears in a “world class journal” begs the question or is an appeal to authority.

    One point of yours with which I agree, though: that the authors of the study are “doing science”. They are. But unfortunately in the modern way. The problem is that the modern way is deeply flawed and a guaranteed producer of wild over-certainty. And I don’t just say this, I prove it. Click here for more information.

    The authors can, in some sense, be forgiven for their egregious errors because their mistakes are everybody’s mistakes. How would they have known they were doing wrong by looking at the example of their peers? Answer: they could not have. Entire fields are fundamentally broken.

  5. Doubting Thomas

    November 9, 2015 at 9:26 am

    Upon further review, I grew more perturbed with your review. You may have been attempting sarcasm, if so please write with more clarity, but since when does science mean unquestionable truth? Someone as bright as you should not spread such profound misinformation. It sometimes bugs me that my business fundamentally uses the same approach as the American legal system: preponderance of the evidence. Maybe in an alternative universe, a better deal has been found.

  6. DT, you forgot the /sarc tag.

    Briggs, you say “Which, scientifically speaking, means this study is crap. And where I use “science” in the old-fashioned, pre-government-grant way. ” And you’re using “crap” in the same way? 😉

    From long observation, and I think most people would agree, moral understanding is rarely sufficiently developed in 5-12-year-olds. That’s why many evangelical denominations reject infant baptism and require what amounts to informed consent before accepting a serious profession of faith. At best this study might identify the morally advanced kids compared to the pack.

  7. Briggs

    November 9, 2015 at 9:36 am


    Attempting? Hmmm. Incidentally, “The Preponderance Of Evidence Criterion Is Absurd“.

    Many people say science means unquestionable truth. Many. If you doubt this, head to your school’s Humanities quad and say, “I don’t believe the threat of global warming is serious.” Report back on your reception. If you survive, head over to Biology and say, “I don’t think neo-Darwianism evolution is an adequate theory to explain human behavior.” You won’t survive this.

  8. I can’t figure out if doubting thomas is a very excellent troll. One look at that graph was enough for me to immediately dismiss any “findings” from this “study”.

  9. Briggs

    November 9, 2015 at 9:42 am


    Tell you what, though. It was good they put that graph in. Probably because they didn’t understand what they were doing, being hypnotized by the knock-you-socks-off weeness of the p-values. This graph is proof of everything I have been saying for years.

    Well, admittedly, it’s not final proof, but it’s close. Final proof would come in trying to use this model to predict how many stickers new kids would share. Anybody want to make a bet with this model?

  10. What a pathetic excuse for an editorial. No attempt to back up your claims. No specific criticisms of the scientific method used. Just another idiot whining when science finds something that doesn’t fit your pre-conceived notions of how the world works. If you want to refute the study, use science. I have no patience for your righteous indignation. It means nothing.

  11. Briggs

    November 9, 2015 at 9:53 am


    Your comment is worthless, because you back none of your claims with any evidence. Whereas I (in the links especially) provide plenty of proof for my claims.

    You can’t just hit-and-run. Come on, then. Tell us exactly where my review is flawed.

  12. Looks like the author’s comment response lends weight to the study. Seems you turn into a bag of douche when confronted with view points at variance with your own. Sarcasm, one-upmanship and a not unsubtle air of intellectual superiority. You my friend are the living proof of the very thing you are trying to refute. I can’t even!

  13. How could they tell which kids decided to give more stickers away when they got home?

  14. There’s a whole bunch of other reasons the conclusions they draw are far-reaching.

    They say the religious kids gave “harsher” punishments, but didn’t reveal how “harsh”. Using the exact same info, we could phrase it as the secular kids were less empathetic to the plight of the victims and less interested in justice. Maybe the religious children have a more developed sense of right and wrong?

    Thanks for your post. I gave a few more on my blog post here:

  15. Briggs

    November 9, 2015 at 10:06 am


    You probably don’t realize your comment applies to yourself, right? Don’t just complain, show us what I have wrong.



  16. This “critique” is so devoid of substance it’s comical. Here’s a list of the criticisms as far as I can find them:

    1) “Yes, this scientifically captures every possible nuance of the scientific concept of altruism, doesn’t it?”

    I doubt the authors claim that. They use a task operationalize a measurable part of altruism.

    2) “The authors asked parents to name their religion. They also asked parents questions like “How often do you experience the ‘divine’ in your everyday life?” They took pseudo-quantified answers from these and combined them scientifically with a quantification of religious attendance and derived a complete scientific quantification of “religiosity.””

    They operationalized religiosity based on how often the kids attended a religious service? I fail to see the problem here. They’re not saying this is a “complete scientific quantification” of the concept. But a useful one.

    3) ” Wee p-values through the terrible abuse of regression on the pseudo-quantified answers.”

    p < .001 is not "wee". Also you don't explain how they "abuse" regression. There are many problems with p-values, but saying any study that uses p-values to assess significance allows you to refute any study you dislike the conclusions of.

    You sir seem to manufacture so many straw-men arguments I'm pretty sure you must own a scarecrow factory.

  17. Briggs

    November 9, 2015 at 10:36 am


    Good for you! Thank you.

    1. “Operationalize a measurable part of altruism.” The question is whether it can be reliably measured at all. I say it cannot. Indeed, it isn’t measured at all. Yet they do claim (which you say they do not) to have measured it. They say, “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.”

    Instead of “altruism”, which they go on and on about, they counted how many stickers were shared in some dicey experiment.

    2. You say, “They operationalized religiosity based on how often the kids attended a religious service? I fail to see the problem here. They’re not saying this is a ‘complete scientific quantification” of the concept. But a useful one.”

    Whether or not it is useful is the question. I say it is not. We have no idea from this study about the “religiosity” of the kids themselves. Did they authors measure the un-measurable “religiosity” of the kids? No, sir, they did not. The quite, quite arbitrarily put numbers to vague and, let’s face it, stupid questions and called it “religiosity.” For instance, I am a Christian. I have no idea what “experience the ‘divine’ in your everyday life” even means. Does that mean I see angels? Do I have to “experience” them daily, or only weekly? What precisely does this mean? Nothing precisely.

    Again I remind you, this measure of religion, bizarre as it was, was applied to kids and not measured on kids.

    3. “p < .001 is not 'wee'. Also you don't explain how they 'abuse' regression. There are many problems with p-values, but saying any study that uses p-values to assess significance allows you to refute any study you dislike the conclusions of." Any p-value less than the magic number is considered wee. I have written a book on the subject. AND I have provided many, many links on how to interpret p-values and regression. My goodness, I can't provide a complete theory of probabilty and a criticism of this silly study all under 800 words.

  18. Briggs

    November 9, 2015 at 11:03 am


    Apropos: Spurious Correlations is now a book! HT Mark Shea.

    Just think to yourself: why do you laugh at these graphs? They pass every statistical test, and give p-values weer than the study above. Yet we laugh. Why? See the links about cause above, but here’s a hint: statistics stinks. Don’t use probability models unless you have to.

  19. Must be liberals studying this–only liberals could go from sharing stickers to altruism . I’ve never heard altruism defined as sharing stickers. I want a study on whether or not sharing stickers relates to anything else in life as far as altruism goes. To actually be worth something, the study should have checked how many kids said they cared and then how many voluntarily shared their lunch with the least popular kid in class for an entire week if told the child forgot their lunch money. This would have to been done publically so all peers knew they were feeding the child now targeted by internet bullies. THAT would have been a test.

    Just to clarify, idiots everywhere take most things literally. It’s part of being an idiot.

    Anon: My neice was trained in school to share everything. She played games in such a way no one won–or more precisely, everyone won. She was taught competition was wrong and winning was wrong.

    Gary: Excellent point. Psychology says under 12 year of age means the concience is not fully developed yet. Bet that’s why we don’t send 5 year olds to jail for stealing a hat. 🙂

    Chris: Interesting comment coming from someone reading the liberal playbook. No actual refutation of the data or method. I’m willing to bet if find a study this poorly done that “proves” what Briggs believes, he will reject it too. Of course, you have to take out “peer-reviewed” because peer reviewers only accept that which agrees with them. Maybe that’s why these studies keep popping up–“science” itself is so biased. (Maybe you could provide a study you like?)

    Phineas: When studying certain ideas and traits, these must be fully explained. The study did NOT measure altruism. It measured how many stickers children would share under a specific set of circumstances. As Rich says, there’s no follow up on what the kids did with the stickers they kept. For all we know, the “less” altuistic ones took the stickers home for their younger siblings. The study measures how well children share in a specific situation. It also measures the differences between children being raised in homes where parents attend church, not the religiousity of the child. To do that, the child would have to answered the questions.

    Doubting: I would be happy to conduct a study to prove anything you want proven and produce the tiny p-values. I get as many opportunities as needed and I get to define the paramenters of the study–ie, I can define sharing, caring, hating, etc. P-values can be produced to back nearly anything if you carefully set up the study. I would love to see this study repeated 100 times in many countries and many religions and see what happens to that precious P value.

  20. 1) The kids are giving away their favorite stickers. With no benefit to themselves. That’s a pretty good, quantifiable, measure of altruism.

    2) “Did they authors measure the un-measurable “religiosity” of the kids? ”

    You’re starting from the premise that a person can’t measure religiosity. So basically you’re saying you can’t study the question scientifically. That’s a pretty lazy way to critique/reject a study. You’ve done that repeatedly about pretty much every topic. Must be nice to have a way out of ever confronting your world view.

    3) I’m assuming you mean you wrote THE book on p values and regressions. You’re pointing readers to those sources as if to say “the methods are bad but i don’t have to show you way you can figure that out for yourself.” I’ve checked out those links. They don’t invalidate anything of regression and do nothing to expand on the many criticisms of null-hypothesis testing. But you can’t provide any alternative methods to show that the results are incorrect. You’ve just decided to call them “bad” when it suits you.

    It’s nice, thought, that in addition to the straw-men you’re adding in the appeal to authority fallacy.

  21. CONSIDER: “Yes, this scientifically captures every possible nuance of the scientific concept of altruism, doesn’t it? Science science science science. Science. It must be science! Scientists wrote this, peer scientists reviewed it, and scientists nod sagely when reading it.”

    CONSIDER also, there were seven (if I quickly counted correctly; or about seven) authors of the study, some of whom probably contributed very little & a minority of which might have only a vague idea of what the study is really about (as is too typical), probably about as many peer-reviewers, probably fewer. …

    … figure about a half-dozen to a dozen people actually involved to any significant degree with this particular study. Even if double that figure, or more, not many individuals.

    But is the analysis & critique about the poor analytical skill coming out of a handful of people?


    It’s a hyper-generalized rant extrapolated to “science.” When the essay author doesn’t like something, its about “science,” not misguided, or mistaken, or conniving/corrupt individuals.
    Logical fallacy: Sweeping Generalization.

    What’s noteworthy is the persistence of the particular pattern exhibited — when scientists, or even just one scientist, mess up the sweeping generalization is invoked to all of science. Whenever possible, “scientism” is invoked.

    In contrast, when a religious figure does or says something profane, or worse…comes silence & focus on something else… Never, ever, a similar extrapolation from the statement/act of one/some to all-in-general.

    What’s really telling in such displays is how, when a study is critiqued, the critique is invariably focused on how bad something was done; and, invariably some obscure example from the lunatic fringe is the focus. Critiques are never offset by an example (never mind exampleS) of a positive study(ies) showing the opposite conclusion per analysis done well. If there’s ever a positive example of how well something was done presented here I can’t recall it — does anyone?

  22. The study is concerned with how children raised in certain households (i.e. Christian/Muslim/Other) performed, not how much the children personally identify with their own religion.

  23. Doubting —


  24. Briggs

    November 9, 2015 at 11:16 am


    1. I say it is, at best, an extremely crude measure of altruism, so crude that it many would say it isn’t altruism at all. It was stickers shared by 5- to 12-year-olds, across many different cultures which may or may not view stickers differently. Do you think the authors would tout headlines and theories about sticker sharing? And in any case, the findings, such as they are and such as presented in their very own graph, are lousy. Not predictive at all.

    2. I am saying you can’t study “religiosity” in the way the authors attempt. And I show why. Did you think I am saying something else? That religiosity cannot be studied at all? Did I say that? Or do I claim, what is true, that attempts to quantify it are silly? Putting numbers to human emotions is widespread, true. But it is discreditable because it leads to vast over-certainty.

    3. You “checked out” the links, eh. That’s pretty fast. I gave hours worth of material which you digested in 10 minutes. Well, prodigies are known. But carping does no good. Say exactly where I have it wrong.

    I’m telling you, and I prove, that probability models can’t ascribe cause. Yet the authors of this study make causal claims. I say exactly what wee p-values mean and don’t mean. I say false dichotomies are involved in any hypothesis test. Show why I’m wrong. Of course, you won’t be able to, because I’m right.

  25. I think important lessons were probably learned from the study. The only people truly benefitting from it are the people doing the study and writing about it. There is hope, that some of the people doing the study will finally figure out the idiocy they are participating in. The only way to truly understand the idiocy is to participate in the process. Altruism is one of those things you don’t study directly. You have to watch when people don’t think they are being watched. As soon as you start measuring, Heisenberg sticks his head in and issues a raspberry.

    We still have to try. There are people out there that can learn well through the writings of others. I am very sure I am not one of them. It is only after I chop a piece of wood to length, glue it up and attempt to fit the mirror in the back that the meaning of “tolerance” smacks me. Not doing such things daily causes me to forget to check for square before letting the glue dry.

    The world keeps turning.

  26. @Ken

    here, apparently, is a study that Briggs likes the conclusions of (and therefore, assumedly, the methods). http://wmbriggs.com/blog/?p=7975

  27. Briggs

    November 9, 2015 at 11:20 am


    Phineas’s humor detector is obviously busted. Briggs said in the link Phineas provided, “This research was in the field of evolutionary psychology, it was peer-reviewed, and the proof entirely statistical, so you know it has to be true. ” He obviously meant that to be taken literally, right Phineas, sugar dear? Sheesh.

    (Incidentally, I’ll be away from the computer for about three or four hours.)

  28. Briggs–

    You missed Doubtimg’s admission: he was blessed (financially?) from having his pseudo-science published. I wonder how much money was lifted from my wallet under the guise of taxation to “bless” him.

    Just sayin’.

  29. Also, Briggs is obfuscating the actual things the authors say. They don’t claim to be looking at the effect of the child’s own religious beliefs; they’re looking at the effect of a religious household on the altruistic behavior. Hence the conclusion:

    “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*.”

    And further:

    “Overall, our findings cast light on the *cultural input* of religion on prosocial 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.”

  30. The kids are being asked to give away scraps of paper with glue on one side. They aren’t giving away their kidney. “Altruism”—is two-sided—it isn’t just the act of something away, but giving something away to someone that needs whatever that something is. Giving a coat to a man who is cold, giving food to the hungry, etc. are examples of altruism. Giving a worthless piece of paper away can’t really be altruism.

  31. Agreed…thanks for the good-humored critique. No temper tantrum from me…crying like a little girl (or for us women) like a little boy. Lol…just pointing out your unscientific reference to gender emotion stereotypes.

  32. So Brigg’s, your entire problem with regression is that “correlation is not causation.”

    And you needed a PhD in stats to arrive at that conclusion?

  33. I’m still trying to figure out how you measure altruism. What are the units of measurement and does the NIST have a standard for altruism?

  34. Ray, altruism is whatever the SJWs say it is. Definitions go against everything they stand for, since then they have to be held to a standard.

  35. “Science means unquestionable truth”. Okay maybe this comment was purposefully snooty, but regardless it immediately shows me you have previously harbored resentment towards science, making me want to discredit the way you analyze a scientific paper….Science is absolutely not unquestionable truth, and every scientist’ head should explode while reading that god awful statement you produced. The subject matter the scientific method produces are data backed by evidence that is taken accepted as what we know until it is proved false. That is the peer review part you always hear about…Well I guess not you, clearly.

  36. Phineas.

    If the p-values obtained from Hypothesis Testing have any usefulness at all it would be to indicate the two variables involved are correlated. As you yourself admit correlation doesn’t by itself show causation. The only hypothesis that could possibly be tested with an Hypothesis Test is one that states X is correlated to Y.

  37. Graham: NO, peer-review is not proof the paper is not false or the theory is not wrong. Peer-review is nothing of the sort. What proves the theory right or wrong is repeatability, correctness of method, etc. Peer-review concerns none of these things. Thousands of bad, later retracted papers, pass peer review. Peer-review picks papers that have what the reviewers consider no glaring errors (math mistakes, etc) and will draw in readers. There have been hundreds of cases of “fixed” peer review with researchers using fake email addresses to get their papers reviewed favorably by friends, computer generated papers were approved with no actual research, etc.

  38. Briggs, it looks like they (you know who) have gone from inserting typos to outright attack. Keep up the good work; you must be making progress.

  39. LOL! I knew you were going to pick up on this story! I don’t like the experiment here very much either, and to be honest, I’m not sure why anyone would do it in the first place.

    I suppose it could be said that religious people tend to be more tribal, and that can have it’s downsides, but tribal also translates to communal, loyal, well-behaved, which can certainly have upsides. I suppose it could also be said that religious people tend to be more uptight, but then that may be more a reflection of class and location. This sort of exercise should give a healthy mature adult a headache.

    It seems misplaced to pin broad behavioral patterns on a single aspect of society. And let’s face it, in the West, in our day-to-day lives, religion is a relatively marginal aspect at that.


  40. Using stickers amongst children 5 to 12 is indeed a poor proxy. My 12 year old son and his friends would tell you that stickers are lame. He would probably give all his stickers to the younger kids. He’s making confirmation this month but his religiosity wouldn’t weigh highly in his decision. Meanwhile, the 6 year old across the street is sticker mad. His bike helmet is covered with them. You’d be hard pressed to pry a sticker from his little hands.

  41. Sheri, Right, but this slew of trolls hasn’t been around long enough to get Briggs’ sarcasm. Especially when he’s in a peevish mood.

  42. Science-y stuff aside, how do you predict change of heart? I can only speak for Christianity but the “fruit” (characteristics that might be associated with altruism which parents are responsible for modeling for their children) are a cultivation of spiritual growth. Christianity is not a guilt ( or innocence) by association model. We (Christian parents) are hoping our children become like Jesus not like us. So, the greatest flaw in the study is that they don’t understand Christianity (which means the same could likely be said for the other religions included) at its core.

  43. This was my first response to the article:

    “I have read the rapport, especially on their definition/determining of religion. The religious denomination was given by the parents who also I have read the rapport, especially on their definition/determining of religion. The religious denomination was given by the parents who also filled in a Duke religiosity test on spirituality and service attendence. The latter doesn’t say much these days, but what I really missed was if the parents were questioned on what their daily ethics, values and norms are based, and how they provide these to their children. I also missed data on specific denomination (how can you compare Catholic natural law ethics to Protestant voluntarist ethics?), social status and how many families participated. The researchers are also clearly focused on attacking what they perceive as general assumptions on religion and ethics. They had many assumptions themselves also as they clearly mistake metaphysical grounding of ethics within theologies for that these ethics are directly religious themselves (e.g. most ethics are not textually based but on some form of natural law). So from a scientist of religion point of view I still have many questions on the way they defined the involved ethical grounding.”

  44. Well done. It is a shame that so many folks can’t nor wish to see past their personal bias. Clearly some of these commentators WANT a study that finds that religious children are meaner than their peers. Thank you for pointing out these problems with yet another “peer reviewed” study.

    There are people that look for truth, even if it is unpleasant or otherwise not fitting with what they’d like to hear.


  45. Phineas: Actuallty, you can’t measurable religiosity scientifically. As was noted, it’s a process, not an end. Also, it’s virtually impossible to define “altruism”. The study’s correct conclusion is “children whose parents self-report going to church and believing in God are less likely to give away stickers to children who have none”. That is the only thing measured here.

    Ken: Actually, sweeping generations are a part of science. When someone announces a study, they mean to convince people that they are right and their theory has value. If you were just studying the mating habits of Brazilian tree frogs, no one would care. Throw in global warming affecting these, and Headline City. You need a sweeping generalization and cause to get noticed. Otherwise, as noted above, the title would have been much different.

    JMJ: I find myself in agree with you. I am afraid. 😉

  46. “First, define altruism. Go on, I’ll wait.”
    And we’re off!

    People’s person definitions of words aside, altruism has a formal definition.
    The belief in or practice of disinterested and selfless concern for the well-being of others. The measurement was to get a sense of the altruism of the participants, not specific aspect of what might be considered atruistic.

    Science does not mean “unquestionable truth” it just means you take a systematic study of something through observation and experiment, like was done in the study.

    Definition for science:
    The intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment

    In regards to the Dictator Game and Altruism:
    Yes, “Dictator game” is a study on Altruism and you agree is peer-reviewed.(http://www.ehbonline.org/article/S1090-5138(06)00093-6/) It is a guide to altruistic tendencies. Now, if you have a problem with peer-reviewed studies, I can’t help you. Peer-review is the best mechanism we have to review studies by having other experts in the field take a look. Who would be a better candidate in determining the validity of study methods? You diminish your credibility by repeating “science” as a way to discredit by insinuating a conspiracy that all scientists agree with other scientists.

    “Religiosity” or “Religiousness” is the based on the “Duke Religiousness Questionnaire (DRQ)… which assesses the frequency of religious attendance rated on a 1–6 scale from never to several times per week (frequency of service attendance and at other religious events), and questions regarding the spirituality of the household (1–5 scale; see DRQ). Average religious frequency and religious spirituality composites were created, standardized, and combined for an average overall religiousness composite.” The study never claim this to be the religiousness of the kids. The point of that part of the study is to correlate the type of religion and intensity of the practices in the households they reside in with the kids altruistic behavior.

    Then you just mention the computer types used in the moral sensitivity.
    You are just pointing out the use of technology when you reference the touchscreen computers…insinuating that the use of technology is scientific when the technology is just the medium with which the study was perform on and with … rather than commenting on the validity of the study used to measure moral sensitivity. Readers can go here to make their own judgments (http://www.ncbi.nlm.nih.gov/pubmed/21616985?dopt=Abstract). Again, a peer-reviewed, useful tool for study.

    The p-value just points to the fact the results are significant and likely to be repeatable. “through the terrible abuse of regression on the pseudo-quantified answers.”, that’s just based on your opinion of the questions and studies used to measure things. You are not providing an unbiased look to the readers by saying “All those dots are the answers to the pseudo-quantifications for each kid.” and not explaining that the “dots”, the points in the scatterplot, are each of the kids results in religiosity of environment, altruism, and age. The grid is a regression plane. I admit, the picture is hard to read since it’s not interactive, but it isn’t hard to see that the data tends toward the the fact that the more the religious a household, the less likely the kids were giving up their stickers to the kids that weren’t getting any stickers by the researcher.

    For this population, and likely for other populations that their findings “contradict the commonsense and popular assumption that children from religious households are more altruistic and kind towards others” is exactly what is shows.. We just need more studies to confirm it.

    As for the idiots writing about it:
    It is a catchy study and fun to write about; Hopefully there will be more studies or different populations to help reinforce or counter these findings.

    As for your links to other blog posts:
    Self-referencing non-peer-reviewed data for factual evidence of your claims… *sigh* and in those, referencing blogs from universities, also not peer reviewed nor studies just more of your opinions.

    In summary, the study is peer-reviewed using some of the accepted tools used to measure data of the nature they were looking at. The correlations are fairly clear look at the bar graphs on this page for a better one to one view rather then a 3d regression plane. (http://www.cell.com/action/showImagesData?pii=S0960-9822%2815%2901167-7). The author of this post has a problem with studies in general and the people performing them and that bleeds through on the attempted analysis of this particular study. The author seems to be offended by the results and should reach out for more studies to be completed to help verify the results… if the author believed in studies usefulness. I would like to know the alternative approach the author would take in place of the tools used to measure the specific data points and why they are better.

    Commenter Note: You’re post was a tantrum in it’s own right. So, please don’t come whining back at me that I didn’t provide facts and proof where you are wrong and that you don’t just have an agenda against what you affectionately call “science” everywhere without better arguments and facts.

    Thank you.

  47. Sander van der Wal

    November 9, 2015 at 2:02 pm

    So, is it now official that Hinduism, Buddhism, Shintoism and so on and so forth are not proper religions?

  48. Using science to clarify the “level of generosity” between “religious”, and “non-religious” kids is like taking shark meat and seeing if a shark wont eat it. Its pretty dumb. Oh wait, I am sorry I didnt use any links to back up my claim click here —-> <——- lol. OMG you people and your proof and oh wait cant use the bible because science cant verify its truth because we cant touch it. Yet, the use of Darwin is the only "theory" being taught in school as complete accuracy of human existence lol. Sad that most scientist even disagree with that theory.

    So lets use stickers to equate the level of generosity between "religious" kid and non, yup thats a great "science" right there. Might as well say kids in Boy scouts who dont pick up trash or help old people cross the roads aren't real Boy Scouts either. Someone just got "scienced"

  49. While I fully appreciate you taking the time to review this, several aspects of your critique make me uneasy. For one, for an individual who has run the academic gauntlet, so to speak, you certainly despise your peers. You’re so familiar with the peer-review process, so why do you rush to bash its fundamental premises? In the second paragraph, we get a healthy dose of sarcasm at “That doesn’t sound like a scientific question, so never mind.” With all of the intense intellectual superiority you boast throughout your review, I’m surprised that you don’t even offer up what you have “proven” in your years of wisdom to be the true meaning of science. If you can frame a question with a hypothesis and test it over and over, are you not seeking the answers to a scientific question?

    Moving on. Your next quandary poses the challenge, “First, define altruism.” The authors define altruism in the only way a biologist can. The donor pays the cost to give a benefit to a receiver, from which nothing is reciprocated. As a statistician, I’m certain that you are familiar with the Prisoner’s dilemma. That is the only definition of “altruism” that a scientist could be using under the scrutiny of their peers. So yes, the general public is not able to wrap their heads around what altruism is to a scientist without that bit of knowledge. But there’s more.

    You then go on to tell us that science means unquestionable truth. Cornell? Really, Cornell pays you for two weeks? No wonder it’s in the summer. No Cornell admin in their right mind would pay you to teach classes during the actual semester, because you haven’t got a clue what science is. Science is questioning. If you can’t ask a question about the results, it isn’t science. That’s the very reason that religion isn’t science; because you can’t question it.

    I’ll grant you all of my skepticism for the Dictator game. The social sciences afford so much to assumption that I can concede that there are holes upon holes in those experimental designs. But they have to publish these studies. It’s a very additive process. Every year they learn a little something new.

    “Yes, this scientifically captures every possible nuance of the scientific concept of altruism, doesn’t it? Science science science science. Science. It must be science! Scientists wrote this, peer scientists reviewed it, and scientists nod sagely when reading it.” Here we have an excerpt of you critiquing the Dictator game. You didn’t bother to enlighten your readers with the “actual” definition of altruism, nor did you grant any acknowledgement of what the biological concept of altruism was, so it really just comes off as more of you not knowing what science is, besides something you hate. More pointless banter. Why the science hate? Did a scientist kick your dog?

    Now, on to defining religiosity. Again, this is no simple task. The researchers did this in the way they decided, for whatever reason, would offer a useful (or significant – hell, I don’t know what their intentions were) insight into how “religious” their sample population was. Once more, I don’t know if you just didn’t get around to taking a philosophy of science class in grad school, or if someone slighted you in a terrible way, but there is not a scientific way to rate precisely how religious someone is. Especially children, who may barely know what you’re talking about when it comes to experiencing the “divine.” One more moot argument from an angry man. Pick at the low hanging fruit all you want, it doesn’t change the fact that they found a small piece of information that you disagree with. And don’t complain about the writer citing to their readers what kind of device they used. You may be writing peer-reviewed papers that don’t have anything to gain from that information, but droves of other researchers can benefit from what is a mundane detail to you. You should try to see the world from a wider scope.

    I won’t argue with your analysis of p-values as pointless. I feel your pain. We need to work harder to stop the rampant obsession with p-values. I highly doubt that the authors wrote this paper with the goal to end religion as a whole in mind. I further doubt that it was their intention to foster disagreement between non-religious and religious families. It feels gross to repeat so much of what previous comments have stated, but you have downplayed the scientific method in such a trite and oversimplified way that I can’t help but identify your misinformation for the few sadists that read all the comments like I did.

    It’s great that you’re active in the world of breaking down bad stats, but maybe you should try to appreciate the role science plays. There are bad scientists, but a lot of us are just trying to make sense of the natural world. While the models might be flawed and the results weak, we have to build our studies on something. There isn’t going to be a masterpiece of a study one day that proves once and for all that god is a lie and Christians are scum. You can’t generalize people like that, so science may never explain the differences between us. Also, you work for the Heartland Institute. You literally get paid to keep the American public stupid and full of fear for a god that was hijacked by cruel intentions centuries ago. If there is an afterlife, you better hope you chose the right side.


  50. Social science doesn’t really count as “science” no matter how many times you use the word in your appraisal of it. It really is more of an art. But that doesn’t mean we can’t find correlations using social science like this.

    What it DOES mean is that we can’t expect to find our data corresponding perfectly to some law, such as the ones we can find in the world of physics or chemistry. Of course, often times we find correlations between completely unrelated things. It is a coincidence, for example, that consumption of margarine per capita in the United States is correlated to the divorce rate in Maine with a correlation coefficient of 0.99 (more than twice the value found in this study). The reason we know this is a coincidence is because there really is no good reason to believe the two things should be related. This assessment is one you just made, subconsciously, if you laughed at the idea of the two being related. It is an accurate one: in order to ascribe meaning to a correlation, we must be able to explain why the two things being measured should be related.

    We have to use common sense to interpret the numbers we find. The “r” value here is interesting. We’ve found a weak correlation between religious parents and selfish children. But can we support it with a logical explanation of why these two things would be correlated?

    Yes, it seems pretty obvious that we can. In every nation this study was conducted in, the predominant religions are religions in possession of a belief that there is a place of eternal punishment where people who aren’t “as good as we are” go to be punished for their sins. It should be obvious that talking to children about this belief is a great way to cause them to feel as if they are better than other people… which may lead them to feel that other people don’t deserve the good things as much as they do. This would easily explain why kids who have religious parents (specifically, kids who have parents who talk to them about religion) would choose to keep things for themselves more often than kids who don’t.

    You’re right though, they should have quantified religiosity better. The correlation would then likely have been stronger.

  51. Confounding error, much?

  52. Wow.
    Besides the poor writing here, I want to make very clear nobody suggested these are black and white in terms of the results and, the researcher plans to expand the study soon as to improve upon it.
    The results are enough though to warrant deeper investigation and a broader study, that’s it.
    Watching people like this squirm over these kind of studies always make me smile.

  53. Thank you for doing your review of the study. I saw links to it on my Facebook feed and figured that there had to be more to the story and this was the only link that came up in the “contrary” column. A quick read-through showed that my suspicions were correct. I can’t speak for the statistics part, but the definitions themselves call into question the value of any statistics on this, even if they showed religious people to be more altruistic.

    Altruism is defined (by Google at least) as “the belief in or practice of disinterested and selfless concern for the well-being of others.” Does giving stickers indicate altruism? Whether that is measurable is irrelevant unless altruism can be defined by giving stickers to someone who doesn’t have them.

    The second part, as has been noted here several times, is that the religiousness of the parents, not the children, is what is measured. And so we have a proxy measurement of religion. That COULD be an accurate proxy. However, as Barna has noted, 65% of children brought up in the (American) church will leave the church within 5 years of leaving home. So is the parent’s religiousness a strong indicator of a child’s? Only if a 65% failure rate is considered a success.

    So, we have a non-consequential definition of altruism (an altruism that doesn’t really cost the child anything) and an inaccurate proxy of the child’s religiousness. It seems to me that the only conclusion that can be drawn here is that children of non-religious parents value stickers less than those with religious parents.

    But as with so much of science today what is missing is “why”. Why do children of non-religious parents value stickers less than those with religious parents? What if the answer is because religious parents use stickers to reward children’s memorization of Bible verses? **Then not only should you count the number of stickers given, but also the value that the sticker has to the individual child as part of the scoring.**

    Geez, the more I think of this study the more amazed I am the grownups will actually take it seriously.

  54. tfnw: Why would academics bash peer review? Maybe because they re the ones who reconize the flaws in the system. Until global warming, peer-review was irrelevent to most of society. Then the global warming crowd elevated peer review to make it equal “truth” so they could squash all disagreement with that which was becoming more and more obviously false in their science. Until it was used as a weapon against those who disagreed with a theory, most of society didn’t even know it existed.

    No, framing a question as an hypothesis and then testing it is not real science. My hypothesis can be that children forced to eat kale turn into mass murderers. I can test it and find out how many come out mass murderers compared to the non-kale eaters, but that does not “prove” anything. I ignored all the other factors in the formation of a serial killer. It has to be a hypothesis with all the parameters correct and the logic correct and the testing must be done over and over to show there is a relationship. One poorly worded study means nothing.

    “Science is questioning. If you can’t ask a question about the results, it isn’t science.” Tell that to the LA Times, the University of Missouri and most any other news media or college.

    “But they have to publish these studies. It’s a very additive process. Every year they learn a little something new.” You’re arguing that addictive processes are desirable?

    Hint: There is no “one” definition of altrusim. It depends on the speaker and the situation. It’s part of ethics, where hard definitions rarely exists.

    If there is no scientifically valid way to rate religiousity, you cannot use the idea in scientific studies. I would assum that was not your point, but thank you for conceding this.

    “While the models might be flawed and the results weak, we have to build our studies on something” That is not what we see. We see science using models and computers AS reality. You are also arguing that a bad result is better than no result. Please defend that idea.

    AND THERE IT IS–the coup de grace: You work for Heartland. Troll City!!!!!

    Anon: You were okay till that “greedy people exist because heaven does”. Your argument could apply to why rich children wouldn’t share just as easily. They came from entitled rich homes. You could also argue the opposite–they share because the have no reason not too. They have all they need. Poor people share in the hopes of getting more. Poor people don’t share because they have so little. The argument doesn’t work. It’s interesting that so many people think Christians believe they are better than everyone else. Why would you think that? Because you believe they’re right? Please explain.

    Brandon: So you’d be happy with studies that proved atheism leads to murders and dictatorships? Atheists are unwilling to share and hate everyone? Interesting. Maybe you can find one and report back?
    Yes, the authors say more study is needed, but wrote the study up so it sounds definitive and the title implies more than the study shows. I have problem with that. It’s also interesting that this study made the news—there are hundreds of similar studies out there reaching various conclusions. Why is this one reported?

  55. I’m not a statistician, so I’m glad that some qualified members of our rational community have refuted your haphazard half-baked attempt to dismiss the study with what I can see as a writer to be uneducated nonsense.

    My expertise is in words and definitions and no, your suggestions for the definition of “altruism” are not anyone’s definition of the word. “No-greater-love, supreme sacrifice, kindness, patience, love, and so on…” Sound like you base your definitions of words on a certain book of nonsense, which may be where you also get your science.

    Altruism: “the belief in or practice of disinterested and selfless concern for the well-being of others.” You should use a dictionary and I’m not being snarky.

    Your definition of “science” is also wrong, and the only “unquestionable truth” in that regard is your incorrect definition of science.

    Science: “the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment.”


    Not that definitions don’t vary slightly, but your definitions are not extreme, but inaccurate.

    I’m glad your artice has been thoroughly refuted, but we all know that whatever fan base you are trying to please will probably not hear rationality or see reality.

    I am a big fan of criticism of studies, which is what science is for, but you sound like an angry Catholic (I’m not accusing you of being one, I’m just saying you sound like one).

    Take care.

  56. Altruism: “the belief in or practice of disinterested and selfless concern for the well-being of others. You should use a dictionary and I’m not being snarky.”This does not mention acting upon that belief. For all we know, the children did believe in “selfless concern for others” but did not view sharing stickers as a way to show that. The study did not address working at food banks, reading to children in a library, visiting a nursing home, etc. The definition used in the study didn’t study a lot of things most people would consider altruistic. Again, your definition also does not include acting upon the beliefs.

    “systematic study of the structure and behavior of the physical and natural world through observation and experiment.” How are ethics part of the physical and natural world?

    It’s interesting that those who say the critique here is “debunked” have done little to debunk it. I have to wonder what the response to this study would have been had the writer been an atheist. Mostly people seem to think that the writer’s religion is the problem. That’s not science—listing why this is valid is science.

  57. Sheri: You wrote:
    The study’s correct conclusion is “children whose parents self-report going to church and believing in God are less likely to give away stickers to children who have none”. That is the only thing measured here.

    You some off: “children whose parents self-report going to church and believing in God are less likely to give away stickers to anonymous children without any stickers and said to be from the same school and similar ethnic group as the child with the stickers.”

    Social interactions are complicated. That’s why they had to include “same school and similar ethnic group” within the study. Sadly, they don’t really allow for religious children possibly having different social circles (maybe church emphasis?) than secular children (maybe school/race emphasis?).

    I won’t say the study is bad, but the conclusions are terrible. Correlation does not mean causation.

  58. “the belief in or practice of disinterested and selfless concern for the well-being of others.”
    This would put everyone who opposes a safety net by the government for the poor, since the money is taken from people against their will. True altruism would only exists where money and time are freely given.

  59. Briggs

    November 9, 2015 at 3:27 pm

    William Strege,

    Boy, I wish you would have numbered your comments. I’ll make a stab. (Your comment went to moderation because it had a large number of links, and I was away from the computer.)

    1. The authors did not measure altruism, but they claimed several times they did. They measured how many stickers some restless kids stuck in an envelope when an adult’s back was turned. Good grief! That’s all this study is. Stickers in envelopes. We know NOTHING about what these kids thought.

    2. “Science does not mean ‘unquestionable truth'”. That so? Nobody says that, do they? Of course many people say this. That is what scientism is; that is what materialism is. I even gave examples in other comments where questions about certain matters will teach you very quickly the definition of sacred. People are so hyper-sensitive they couldn’t see that this comment was meant as sarcasm? Good googly moogly.

    3. Dictator game and peer review. The hell with peer review. Peer review is a boondoggle and of no use in its present form. It produces and enforces mediocrity and, like in this study, ensures the persistence of error. That you “can’t help” the problem of peer review is not interesting. Just tell me what fallacy it is that says a study (or method) is correct because it was peer reviewed?

    4. “Religiosity”. Of course the authors wanted us to infer that religious kids, kids brought up in religious households, were different than those brought up in what they call (but who knows the truth) non-religious households. Now everybody knows the religious attitudes of 5 and 12 year-olds are worlds apart, and recall we know NOTHING about why kids put stickers in envelopes. Plus, measuring “religiosity”, I mean putting a number to it, is asinine. Is a person with a religiosity of (say) 4 twice as religious as a person with a religiosity of 2? And is a person with a religiosity of (say) 2 twice as religious as a person with a religiosity of 1? It is rank pseudo-science to suppose we can put a number to a person’s religiosity. Religion is hideously complex, multi-dimensional, fluid, and difficult. How much do you agree with that statement on a scale from -17.4 to 38.3 in increments of 5/6? This kind of thing is commonly done, it’s true. And that’s because peer review ensconces the fallacy.

    5. Computer type. I teased them about their language because it was unnecessary and made their study sound so sciency. Did the people in the study wear white lab coats? Incidentally, I typed these comments on a Dell Inspirion 3000 running Kubuntu 14.04.3 LTS with a standard QWERTY keyboard in a room illuminated with 29.8 lumens per square foot.

    6. “The p-value just points to the fact the results are significant and likely to be repeatable.” Brother, on this subject you are ignorant; a vast ignorance. I mean this word in its technical sense. You have obviously made no attempt to read the links I provided.

    7. You say “For this population, and likely for other populations that their findings ‘contradict the commonsense and popular assumption that children from religious households are more altruistic and kind towards others’ is exactly what is shows.” No, sir, it does not. Altruism was not measured, and the quantification of “religious households” is stupid, as I have shown. How often you Bill (if I may call you Bill) experience the divine in daily life? And we haven’t even yet talked about how Muslims were the largest segment, how so-called “non-religious”…oh, the heck with it. The authors of the study cannot back their claim. Even their own picture, which I show, proves this. Sheesh.

    8. “It is a catchy study and fun to write about”. True. So much fun that the people writing about it never bother to learn what they’re talking about. Hence, idiots.

    9. “Self-referencing non-peer-reviewed data for factual evidence of your claims…*sigh* and in those, referencing blogs from universities, also not peer reviewed nor studies just more of your opinions.” Did you remember the name of the fallacy from (3) above? This comment of yours is interesting. You think it saves you from answering my charges (about statistics and so forth) because it doesn’t appear in some official venue. Don’t take this wrong, but that’s pathetic.

    10. “In summary, the study is peer-reviewed…” Boy do people have a hang up about peer review. Truly, scientists are our modern priests, eh?, eh?, where the ritual of peer review guarantees purity. Dear Lord.

    11. “You’re post was a tantrum in it’s own right. So, please don’t come whining back at me that I didn’t provide facts and proof where you are wrong and that you don’t…” You did try to answer, and that’s to your credit. But your claims don’t hold up. Mine do, though.

  60. Briggs

    November 9, 2015 at 3:33 pm


    While I away, several people tried to comment with profanity-laced criticisms. These were caught by the filter and deleted.


    These are the best insults you can come up with? 900 words and not one laugh line. Do you work for NPR?

  61. Just Another Christian

    November 9, 2015 at 3:33 pm

    So… I understand where you are coming from. I get your point and am not refuting your conclusion (though if you shored up your writing maybe you wouldn’t have so many people writing you off). My problem is this (and I am going to assume for the moment that you are a Christian): as people of faith we are asked to be understanding, forgiving, and caring; we are supposed to look at how we can be better people before we point our fingers at others. Even writing this makes me think I am calling you out on things while I should be looking at myself instead but I thought this was important enough.

    In your article you are exceedingly rude to people who may disagree with your opinion. When Jesus was presented with immoral, deceitful, callous, hateful people he did not spit on them. He washed their feet, asked them how he could help them and gave them advice to make their lives better (and yes, I included his own disciples in there on purpose). He loses his cool only once, and that was over a lot more than a study saying religious people are not as nice as other people.

    Now lets look at your article. I grabbed a few quotes but I am sure there are other applicable ones to be found. “Asinine studies” – rude and demeaning attack. “…throw a temper tantrum and (for you men) cry like a little girl…” – again rude and demeaning but now you are also being sexist in more ways than one.
    “keep your mouth shut” – rude, aggressive and boarder line threatening; readers are supposed to keep their mouth shut or what? “I mean, it stinks to high heaven” – this one makes me cringe as you use religious imagery to dismissively condescend to people about how religious people are just as kind and caring. “I’m too tired to make a joke about science” – why are we making jokes about science? You would not be on the internet, much less have a computer or even electricity if it were not for science. Obviously you have nothing against science and are more frustrated with the scientists – flawed humans like you and me. “means this study is crap” – no description needed. “My goodness! How scientific!” – again with science bashing. Also the sarcasm is uncalled for, it sounds petty. And finally the best of them all:
    “I am too tired and too surly this morning to put this more politely.” So does that mean we can all be rude, short, and condescending when we get tired?

    I think my point is clear. More importantly if you had cut out many many of these hurtful comments your article would actually have the effect you surely intended – for people to understand that the experiment was flawed and we should not simply say something is true because one scientific study says so. We should question, read, and study to come to a better understanding – something both science and religion preach.

    So instead your article had the exact opposite effect – it showed people a presumed religious person not only being rude, disrespectful, not understanding but also being a hypocrite. Instead why not try to be understanding, accepting, open to others and their (perhaps misguided) understanding and help them become the best version of themselves. If you act like this in your everyday life you will only promote a poor image of religious people; and this will only continue as children use your behavior as a model for themselves. Honestly, as a christian I was more appalled by your response than by their blunder.

  62. <irony> Briggs, include more of these. It may help.

  63. Briggs

    November 9, 2015 at 3:47 pm

    Just Another Christian,

    Me? Bash Science? Heaven forfend! What is more wonderful than Science! There is no bigger champion of the majesty and sublimity of…why, just last week I came across a new kind of toothpick that does two teeth at once and automatically updates your iPhone with the number of particles removed so that you can track your dental hygiene through time. If I was a little richer I would have bought the version that had a small camera so that I could have uploaded the videos to YouTube.

    And you called me a sexist! You hurt my feelings. Let me know how you’re going to punish yourself.

  64. Just Another Christian

    November 9, 2015 at 4:11 pm

    Sheri: First I would hope we all understand that correlation is not causation. But that does not mean you cannot prove causation. Most of your disagreements with the other commenters hinges on this factor.

    Second you cannot isolate one factor in any population so instead researchers are supposed to pull from a wide range of types of people so to counterbalance those influences. There is literally no other way to work around this problem aside from raising cloned children in controlled environments.

    Third I am surprised you are still disputing global warming, because even if it were a hoax we should learn to treat the earth with more respect so why even argue the point?

    Fourth you cannot prove religiousness but you can prove other factors such as amount of time spent in religious activities, amount of time spent reading religious texts, amount of time listening to religious music. From the sounds of it you are trying to destroy the entire science of psychology – because religion is definitely part of each persons psychological makeup and you can have experiments to understand the human mind.

    Fifth science does not work against the idea of God – it simply examines his work more closely and less biasly. We can learn from science when reading the Bible so to not read only what we want to but to ready earnestly so to understand earnestly. So why should science not be able to shed some light on religion? You seem to look for fundamental issues with science and its ability to even in a perfect scenario explain the impact of religion on young minds.

    Instead of dismissing all opposed why dont we work with science to raise our children better? What if some shred of truth exists in the study, what if on all levels, what if God himself would say that religious parents are raising their children to be less understanding than nonreligious parents? Why do we automatically become defensive? Instead we should look to become better not matter what faces us. Because how sad would it be that we become worse parents simply because we were more focused on how science was wrong about us and less on actually practicing what we preach?

  65. Just Another Christian

    November 9, 2015 at 4:28 pm

    Briggs: Why are you so bent on being hateful? I ask you to be a little kinder and you throw this back? You are only proving me (and sadly enough the experiment you detest so much) right.

    If you hate science so much at least practice what you preach and go live in the wilderness or with the Amish and experience it for once. Until then the sarcasm is just a front you’re using to avoid giving any real answers.

    I called you sexist because you were being sexist. So what am I going to do because I hurt your feelings? Tell you again to stop being sexist. It doesn’t look good on you.

    (Also sarcasm doesnt make you right, start using facts or start losing the respect of anyone listening.)

  66. Briggs

    November 9, 2015 at 4:30 pm


    From the author of the study: “My guess is they’re [Christians, mainly] just going to deny what I did — they don’t want science, they don’t believe in evolution…” Et cetera, et cetera.

    It’s Science!

    Just Another Christian,

    I’m also a racist. I’m hurt you didn’t see that. What is it with you and wanting to hurt my feelings? What have I ever done to you? You’re nothing but a bit meanie-pants.

  67. Just Another Christian

    November 9, 2015 at 4:38 pm

    I give up. If anyone else wants to try be my guess. Briggs – I was actually on your side, don’t know what must have possessed me. Maybe it was science, or sexism. Not probably the racism. Im sure you will have some fun responding – have a nice life. I hope it proves to be educational.

  68. Paul: You’re right. The title you give is much more accurate.

    Whomever it concerns: People most certainly do say science is the unquestionable truth. Check out global warming sites if you doubt this. Science is so unquestionable that no dissenting views are allowed. None–disagree and you’re toast.

    Just another: Really–the WWJD? In reality, we have a few speeches that were made by Christ. We don’t know if he was always PC or not. I’m sure he would have been PC, however, because reality is just too harsh on people. Perhaps you’re right. No, not really. (This reminds me of the person who went onto a blog, took the writer to task for his profanity, put 30 links to her own writing, informed him she reproduced his posts on her blog, minus the profanity and figured he would be estatic that she “shared” his writings and cleaned them up to her standard. He was not and was very clear in letting her know.)

    It is this PC behaviour that just resulted in the University of Missouri president resigning because a few students were angry that they were not being treating right. It’s why Yale students are afraid of Halloween costumes and demand they be protected. Nicey-nicey is why this is happening and you are certainly aiding and abbetiing the PC crowd.

    How do you prove causation? In social science, the only measure is the wee P-value. Wee P-values show correlation. How do you prove causality in human behaviours? If you know the answer, you will win a nobel prize and be called blessed, by both dictators and idealists. No one can prove causality in human behaviour. There are too many factors. In fact, that is actually the reason religion exists–God explains this as original sin and our nature. Science cannot provide the answer.

    Why argue global warming? Because it’s not about treating the earth well–it’s about income distribution. If what I do causes extreme hardship but I meant well and the earth was treated well, that’s all that counts?
    Yes, you can have experiments to understand parts of human behaviour. However, they must be set up carefully and the conclusions will always be very limited in scope. If you cant prove religiousness, then the studies should not use the word. Yet they do.

    Science can shed some understanding on religion but the Bible clearly says belief is a matter of faith. God gave no scientific proof of his evidence and it seems that was on purpose.

    I am all for raising our children better–starting with understanding the truth, propaganda, how science works, how morals work, etc. Again, you’re okay with a shred of truth amoung thousands of lies? How does that work?? Where did God say “Go ahead support a lie if it helps” “Go ahead and preach lies if people feel better?” I am appalled that a Christian is in favor of lying to children. Obviously, you do believe the idea that it is more important to be liked than right. That is NOT what God said. Anywhere.

    Since Just gave up, I’m posting this for anyone who might be interested.

  69. Ye Olde Statistician

    November 9, 2015 at 4:51 pm

    Its authors certainly are doing science,

    I thought they were doing psychology or perhaps sociology.

    The results are enough though to warrant deeper investigation and a broader study, that’s it.

    This is the Universal Conclusion of all funded studies. More funding is needed.

    saying [that] any study that uses p-values to assess significance allows you to refute any study you dislike the conclusions of.

    It would be well to understand the nature of the criticism. Examine again the graph reproduced at the top. My first thought on seeing it was: “Nothing to see here; move along.” That is, the model is not skillful. Look at the spread of data points around the regression plane! There is no tendency of the points to cluster around the plane. This is like the notion that the average human being has slightly less than one testicle. It’s a numerical average, but has no physical meaning.

    The wee p-value referred only to the hypothesis that the slope of the regression plane is 0. The slopes of regression lines are seldom zero, so a significance test is hauled in to ask whether the observed slope(s) could be found in a random sample if the actual slope of the model is zero. The answer seems to be “no.” But what is the physical meaning of this? (Here is where the voodoo sciences part company from the hard sciences like physics, chemistry, and even biology.) One conclusion might be that the samples were not random (and therefore the significance level is compromised). Another might be that children who don’t share stickers feel guilty and turn to religiosity. Or perhaps there is a lurking variable Z which is causitive to both X and Y. Or, like the % imported cars in the USA vs. % women in the workforce, there is no causal link at all, only coincidence.

    Even in the hard sciences, engineering, and the like, we often had to resort to surrogate measures: hardness of steel instead of tensile strength, viscosity instead of degree of polymerization, radiation backscatter instead of density of coal, etc. However, both the desired measure and the surrogate were well defined and a solid correlation and causation were known. Hence, you could measure the backscatter on a bunker of coal at a power plant and estimate the density of the coal in the bunker — and propagate the variance through the transfer functions! (Folks often forget to include the variance on these correlations.)

    And we will not even speak of the distinction between testing the parameters of a model and the actual real world measurements. The wee p-value refers only to the former.

    children forced to eat kale turn into mass murderers. … ignored all the other factors in the formation of a serial killer.

    Sheri! This conflates mass murderers with serial killers! The former would more accurately be “parallel killers.” (Just noting how important definitions can be.)

    Of course, Sheri’s main point is the doozy. Between any two groups of human beings there will be differences and there is no reason to suppose that the label used for the two groups is the only or even the most important one. For example, in troubleshooting a problem involving battery life, the late, great Ellis Ott found a statistically significant difference between the batteries processed on a station using fresh hydroxide and those processed on a station using recycled hydroxide. Ott pointed out that there was no reason to suppose that it was the nature of the hydroxide that accounted for the difference, only that there was probably something different between the two stations. (It might for example have been operator skill.)

    Another, more public example, found a correlation between the time women spent working in front of CRT screens (remember them?) — i.e., at computer terminals — and cancer. Much hoo-hoo was made; but it turned out that the women who did such work (“pink-collar” workers) also smoked more, ate fattier diets, exercised less, etc. compared to the women who did not (lawyers, managers, doctors). That is, the difference in effect was not demonstrably due to the label used to form the categories.

    the definition of “altruism”

    I must be getting old. I can remember when the evolutionary biologists claimed there was no such thing.

    you sound like an angry Catholic

    You say that like it’s a bad thing. Probably blowback from a Guardian reader.

  70. Matt,

    I’m hurt that you’re SO mean !!!!

    Plus!! you were not respectful of REAL scientists who are just trying to find out THE TRUTH !!!!

    Plus where do you get off MAKING FUN OF genuine science things like p-values and Likert Scales ?!!! EVERYBODY uses these !!!!! For EVERYTHING !!!!!!

    PLUS !!! “Everybody” “knows” that “religious” 5 year olds are WAY more altruistic than sensible 5 year olds — EXCEPT — IT’S NOT TRUE !!!! Ha Ha about that !!! This is why we NEED SCIENCE !!!!! And you are AGAINST Science !!!!!!! Because you LIKE the WRONG ANSWER !!!!!

    I’m feeling really, really triggered right now. I really want you to realize this !!!!!!!

    PLUS !!! you said NOTHING about Standard Quantized Singing Units !!!!

    And you SHOULD !!!!!!

    Since EVERYTHING can be measured by Standard Quantized Singing Units !!!!!!!!!!

  71. Your affiliation with the Heartland Institute makes you an inherently biased source to be commenting on this. Your personal beliefs, and perhaps the source of your paychecks, are dependent on this study being false. You wouldn’t be qualified to sit on a jury for a case like this because you couldn’t be fair. Your condescension ruins anything that made this article viable.

  72. “These are the best insults you can come up with? 900 words and not one laugh line. Do you work for NPR?”

    This is why I keep coming back here.

  73. Briggs

    November 9, 2015 at 5:03 pm


    Fascinating comment. Stupid, wrong, and ill-informed. But fascinating. If you can tell me the fallacy that says that because I gave two speeches at Heartland over the entirety of my life that therefore my criticisms of this silly “altruism” study are wrong I will give you a crisp new dollar bill. You have five minutes before this offer expires.

    Just Another Christian,

    So it’s goodbye forever, then?

  74. Just Another Christian

    November 9, 2015 at 5:03 pm

    Sheri: I hadn’t given up on you but I guess I will. No where did I say to give into lies, my problem was entirely on how we react to things that we assume to be lies just because we dont agree with them. If we cannot prove religiousness we can certainly not disprove such a causation either (and please tear up my proving a negative point, I know you want to). Because at the end of the day, yes God can understand all of this and yes religious (or more specifically Christian parents) could be raising their children in worse ways. We will only know once we die. But when I go I would much rather be asked why I gave heed to such scientific trash and in doing so attempted to raise my children to be more kind and understanding than to be found guilty of willful ignorance and blind disagreement and in doing so continued raising my children to be less like Christ. Choose your poison, I certainly chose mine.

  75. The technical scientific issue here is the distinction between reliability and validity. Briggs and others are pointedly questioning the VALIDITY: that is to say, there is an argument to be made that what these numbers are measuring is NOT what the study’s authors claim to be measured. This is an essential and reasonable argument to make about any study. Most likely, even the study’s authors would admit the difficulties of, for example, labeling giving away stickers as “altruism”, that it is at best an approximation of altruism. (This is the case with most papers/presentations in the scientific community. Part of the discussion is “how strong” is the case being made, where are the weak points of the study.)

    Another technical question is what is the RELIABILITY of the methodology of the study. A single paper is unable to determine whether a study is reliable. A study’s methodology is reliable if it is repeatable, if you get the same numerical results with different authors, different children, different families. The reliability of this study has not been determined, because no one has attempted to repeat it, yet.

    It is possible for a study to be reliable, but not valid: whatever is being measured can be measured consistently under varying conditions, but it isn’t really measuring what is claimed to be measured. It is possible for a study to be valid, but not reliable: it is measuring what it is claiming to measure, but the measurement is so difficult to do correctly, that one cannot glean meaningful results. A good metaphor is the old joke about the drunk who lost his keys in the dark corner of an alley (looking there would be “valid”), but instead is looking for them under the street light because it’s easier to see (looking there is “reliable”).

    My personal opinion is that this study is invalid on epistemiological grounds: mapping “giving away stickers” to an overall moral attribute of “altruism” is a highly questionable assertion. It would be better to have multiple measures of altruism, or at least have a separate study to confirm the validity of the “giving away stickers = altruism” assertion. The “religiosity” measure would appear to be somewhat more valid – it’s fairly easy to measure attending worship services, etc., that would at least intuitively correspond to what most people mean by being religious.

    I would ask readers to look at that graph at the top of the page. That graph is the entirety of the results. Look at those points: they’re all over the place. When you have points like this that are necessarily in a box, and they’re all over the place, you’re not measuring lines or trends. You’re measuring which parts of the box don’t have quite so many points as other parts. You need to be very careful before assigning MEANING to that absence of points.

    In this case, the sparse areas are the upper right and upper rear areas. It’s that absence of “10s” that give the -.173 correlation trend “line”. The corollary of this is that older, less religious students are more likely to just give away all of the stickers. By what reasoning is “give away all the stickers” more “altruistic” than “give away half the stickers”? How do we know that the reasoning isn’t “I don’t care about stickers?” Might a better hypothesis be that the 10-12 yr old students think that stickers are for 1st graders and kindergarteners? (The age trend is larger than the religious trend, which is very slight.) Another hypothesis is that religious kids are more likely to “follow directions”, and keep some and give away some.

    See where I’m going with this? If we take away the assumption of “altruism”, and just talk about stickers, we can come up with a dozen different hypotheses for the observed pattern. The p-value only says, “There is a pattern here.” It doesn’t say what the pattern is beyond mathematically quantifying it.

  76. Just Another Christian

    November 9, 2015 at 5:10 pm

    Briggs: Yes, but it was fun while it lasted. (Just for you 😉 I figured I couldn’t go without leaving a little bit of sarcasm. It’s the new rage, everyone is doing it.)

  77. Sander van der Wal

    November 9, 2015 at 5:17 pm

    Regarding altruism, there are a couple of scientific theories defining and explaining it: https://en.wikipedia.org/wiki/Altruism#Scientific_viewpoints.

    So, instead of looking at a dictionary, a scientific paper has to state which scientific theories of altruism they are using and testing against.

  78. I am sorry, I didn’t read your post in detail. There were way too many errors at the top that jumped out. First is your assumption that behavior can not be part of science. There is a scientific field devoted to behavior. It may not look exactly like physics or exactly like chemistry but it does use the principles of science and applies it with success to the field of behavior.

    In this case their measures of altruism and religiosity are clearly defined and measure what they seek to measure. The study shows that, using these measurements, kids raised by religious parents are less altruistic. It’s pretty clearly stated. It just sounds like you don’t like what this paper says so you’re going to randomly yell the word science a lot. You never actually bring up proper arguments for why the game doesn’t measure altruism or why the interview doesn’t measure religiosity.

    I’m sorry, but science isn’t about making sure your feelings aren’t hurt. If you can’t handle things like this another field of study might be better suited for you.

  79. Forget how about how it was defined; “Altruism was calculated as the number of stickers shared out of 10.” Is there not a problem with that?

  80. This review is spot on. Did anyone one mention that at least 2 of the authors of this study are atheists? One only need to look up their Twitter feed, if they haven’t taken down their Twitter feed out of shameful bias yet. Also, 33 kids from Chicago represents the USA? Was it the part of Chicago where the stickers resemble US FOOD STAMPS? Were the stickers ‘My little Pony” themed and offered to atheist boys- in which case, give them to the girls. Gender, Socio-economic status and paltry sample across ‘the world’ all are contributing factors of this ‘study’ being crap. Also, it helps if you have children, to understand children.

  81. Gabriela: So no one is qualified to sit a jury because everyone has prejudiced beliefs. I’m good with that. Did I mention the new IPCC chairman once worked for Exxon?

    Just: This was not disagreement because we “don’t agree”. It was bad science. No need to address your negativity–oops, your proving a negative point. You know the problems with that. You have every right to choose whatever poison you want. Go for it.

    Daniel: There are many comments here explaining why the paper is not valid. You seem to ignore all of them. Psychology is questionable as a science. It’s certainly not “hard” science with yes/no answers. I will grant you that the psychology of marketing is very good, though it’s based on much criteria that are easier to define and measure. Does commercial A lead to more buying or not? Beyond that, it’s really more of an art than a science. (How does psychology explain only reading part of an article and then commenting on it?)

    Katie: Apparently not to the study’s author, but to the rest of us, yeah.

  82. Keeping this short. If you’re just going to throw out accepted methods of studying populations because you feel its wrong without any real evidence and no alternatives, there is no point in a discussion. Your comments illuminate the fact you are more interested in trolling commenters. The responses to your comments are in my original statement and like the study, you can go back and read more carefully and critically. Good luck and I hope you someday understand where you went wrong. Good day.

  83. C’mon! We can do it!
    Break number of comments on a Briggs post record!

    What is it about this sort of thing that draws so many comments (including mine) when the more interesting Summa Contra Moderna series does not?

  84. Elostirion: My guess is this is far easier to understand and relate to (said by one who is not a fan of Thomas)

    Just for Fun since we’re in deep there, I looked up headlines on this study:

    CHILDREN who are raised in religious families turn out to be meaner than their parents, a new international study has claimed. (news.com.au)

    How religion makes us meaner

    Does growing up in a religious family make you MEAN? Christian and Muslim children found to be less altruistic than the offspring of atheists

    Religious Children Not as Kind as Their Secular Peers, Study Shows

    Study: Religious Children Meaner, More Punitive
    (opednew.com–liberal site)

    and my personal favorite:
    Study: Religious Kids Are Jerks

  85. Ye Olde Statistician

    November 9, 2015 at 6:35 pm

    there is an argument to be made that what these numbers are measuring is NOT what the study’s authors claim to be measured.

    Heh. My old boss, Ed Schrock, used to call measuring the wrong thing “errors of the third kind.”

    A study’s methodology is reliable if it is repeatable

    And so few of them are that those who try are called “replication bullies.” Apparently, it hurts people’s feelings to learn that their “significant results” can’t be replicated!
    Even in relatively hard areas like medical research, we find replication failure in as many as 80% of the papers. However, these failures seldom make the evening news and the public is left only with the impression made by the original announcement.

    It may not look exactly like physics or exactly like chemistry

    Got that right.

    The Scientific Revolution was developed in response to problems in the physics of local motion. The further we get from that situation — inanimate objects, identical and identically disposed, subject only to outside forces — the less well the methods apply. Even physics has no analytical solution to the Moon’s orbit, the original three-body problem. When the number of bodies is greater, the scientific method founders, loses its mathematical rudder, and must fall back on statistics. But statistics is useful only when the many bodies are essentially uniform, as in thermodynamics or quantum theory (“disorganized complexity”). When the bodies are not uniform or when they are related to one another in various ways (“organized complexity”), you cannot replace the individual units with an average unit and get useful results. The manner in which the units are interconnected must be taken into account.

  86. Funny how this is the “go-to” study, when every other study showing charitable giving shows just the opposite.
    Here’s just one: https://philanthropy.com/article/Religious-Americans-Give-More/153973

  87. John M. Christensen

    November 9, 2015 at 7:15 pm

    So, did the other kids actually get those stickers after the researchers counted how many had been put in the envelope?

  88. John M. Christensen: Good question! Would that then measure the researcher’s altruism?

  89. Jessie Hudson bleats:

    “This review is spot on. Did anyone one mention that at least 2 of the authors of this study are atheists?”

    Uh oh. Here comes Dr. Briggs to ask you if you know the name of the fallacy you just committed. Or, maybe not.

  90. Re: Spurious correlations

    What is spurious about correlating Arcade Revenues with Computer Science degrees – MY GOD! – how is that spurious????

    (did they correlate comic sales as well?)

  91. Why did the researchers limit the range of altruistic behavior practiced to sharing stickers? I’m not certain about Islam, but Christianity largely teaches people to be charitable when a stranger needs food, clothing, shelter, the sorts of things that can mean life or death, sickness or health, not superfluous fluff. They are also encouraged to provide spiritual support for the lost, and time and compassion for the unfortunate, ill and dying. How does the sharing of stickers correlate to the actual practice of more meaningful forms of charity, those actually encouraged by Christianity? Why did they not collect data on time volunteered helping an elderly neighbor, sharing food with a hungry schoolmate, or standing up to a bully? These would be more meaningful measure of how much Christian values might impact a child’s social behavior.

  92. James said: Might a better hypothesis be that the 10-12 yr old students think that stickers are for 1st graders and kindergarteners? (The age trend is larger than the religious trend, which is very slight.)

    Agreed, I made the same point above anecdotally using my twelve year old son and my six year neighbour as an example.

  93. JohnK, I’m sending you my cleaning bill. I spilled red wine allover the sofa.

  94. “Might a better hypothesis be that the 10-12 yr old students think that stickers are for 1st graders and kindergarteners? (The age trend is larger than the religious trend, which is very slight.)”

    Which suggests that the obvious follow-up study is to create an equation mapping lack of religious upbringing to dismissiveness towards stickers.

  95. Top stuff, Briggs; if you’d used a more emollient tone you’d never have drawn all these atheist fundamentalists out of the woodwork. And I’m an atheist, who happens to value christian morality almost as much as intellectual honesty.

  96. This study is reminiscent of those studies that discovered that Democrat voters have a higher IQ. They both demonstrate that some adults should never be let loose near a bunch of numbers.

  97. I am commenting on two fundamental flaws that the “experiment” has. I think Briggs says some of these things in a roundabout way.

    1) The definitions of altruistic and religiosity are suspect. Because they do not us previously defined method of measurement of these two attributes (Ray has it right), the study seems to make up its own. This makes for difficulty in quantifying altruism and religiosity between different studies, or even difficulty of the reader understanding what the study means by these two terms, and maybe the researchers don’t know, either. Indeed, they seem to consider only attendance of organized congregational meetings, but atheists can be very religious. Atheists have a deep belief that there is no god(s), and they often worship this belief – sometimes to the point of insisting that others should share their belief (if you believe in God, you are evil, but if you don’t then you aren’t). This makes them likewise religious. Yet the measurement of religiosity would miss these believers, as they do not have an organized method of worship, but they can, do, and will reinforce their beliefs among themselves, sometimes, as they interact with other atheists.

    2) How their definitions were created is in doubt. It is unclear whether they measure what they intend to measure. There is an assumption that all people want the same things, so if one kid wants a certain set of 10 stickers, then the assumption is that all kids would want the same set, thus the only altruistic option is to give up the stickers the kid has chosen. My experience, though, is that kids by themselves select different toys than other kids who are by themselves; different people have different desires, so keeping the desirable stickers can be just as altruistic, because the kid would not be imposing his own desires and beliefs upon someone else. However, most people assume that because they want something themselves then everyone else wants the same thing, and the researchers seem to be relying upon that assumption as the only determination of altruism.

    The researchers may be tempted to adjust their definitions of altruism and religiosity while working with their data in order to find results that produce wee p-values — thus are publishable — and this risks inserting an unintentional bias that could affect their results. When searching for a wee p-value, they may settle upon the first conclusion that gives them such a value, and their search will drift in the direction of any pre-conceived notions that they have. The researchers may have asked questions about charitable deeds, but found that they did not produce the needed p-values for publication. We do not know.

    Having worked in test engineering for a couple of decades, I know the importance of designing an experiment that is not only able to collect data that proves the question (e.g. the unit under test works; even the definition of “works” has to be clear) but also to disprove it (e.g. it does not work), but if experimenters are allowed to massage their data or definitions in order to find publishable results, then it is difficult to know whether they have actually proved anything, including what they say they proved.

    Unfortunately, the researchers’ “experiment” does not have sufficient definition to mean much of anything. It is like the classic IQ test: it measures whatever it is that the IQ test measures. This experiment measures something about sharing stickers, but what it measures is unclear. Perhaps it measures the researchers’ haplessness to choose 30 stickers that kids do not like enough to want to keep, or (altruistically) don’t want to even inflict upon some poor unknown other kid and would rather take home to throw away. We really do not know. Nor do we know whether any of the kids is an atheist who is forced, kicking and screaming, to go to temple every week or whether he is very pious but cannot convince his parents to take him to temple at all. This meaningless “experiment” is poorly designed, but the peer reviewers failed to notice — perhaps because they also do not know how to design a meaningful experiment. There is a reason that psychology is considered a “soft” science — the conclusions are so rarely conclusive (because different people behave so differently, as demonstrated by reading these comments).

    Finally, I would like to emphasize that, misused, statistics can be deceiving. For instance, the “average American” is a hermaphrodite, but very few actually are. As with modeling an “average” American, the plot from the “experiment” is a poor predictor of any individual, and no “law” of nature has been found. Indeed, the researchers drew a conclusion from “shotgun” data that actually demonstrates that there is no real relationship between the three plotted attributes (once again, the peer reviewers failed to point out this problem prior to publication). It seems that many commenters to this posting do not realize that it is all too easy to lie with statistics, even unintentionally, and especially when assigning causation from an assumed correlation (no correlation can really be seen in the data, demonstrating that the published 0.408 value is worth very little if we are to predict any individual; the researchers seem to be only interested in the statistical numbers and have ignored the actual data — remember: the data is the data, the statistics is not the data; at best the statistics can tell us something about the uncertainty concerning the data). Remember, statistics does not show cause. One commenter’s definition of science would have made Aristotle seem scientific, yet several of his deductions, several of his methods, and his hubris set back science and human advancement a couple of millenniums. Another commenter’s definition was that because there was observation and experimentation it must be scientific, but a poorly designed experiment is not science. It appears that people who know little about science and have forgotten much of what they learned in high school are trying to tell the rest of us what science is.

  98. This caught my eye “Science means unquestionable truth.” This actually makes more sense to me when applied to some organized religions. But the point of science is to change, to always be questioned. As our understanding changes, so too does science. We test it again and again, with the intention of altering our previously perceived observations.

    It sounds like the author is saying this study was 100% false. It was A study that shed light on SOMETHING. Statistics are convenient, but they never show us the whole picture. I think that’s what people forget. There’s always the exception, and those individuals shouldn’t be ignored. So it isn’t right, it isn’t wrong, it’s a study that gave us some collection of results and that was outlined so that someone else later (several people for that matter) could come along and duplicate it but with altered methods. It’s also not a simple 1+1 thing, so as far as I can say, it likely will never be completely right or wrong.

  99. Elostirion,
    OK. I’ll add to the hit count.

    Some one up above.
    Best point yet. If you cannot give me measurement units for altruism, you are not doing science.

    Someone else up above.
    Yeah. I am just repeating what you said about what is the altruism scaling factor. 5 stickers might be better than 2 stickers but 10 stickers translates to “I don’t give a shit about what this adult is doing”

    All of you progressive weenies.
    Adding “ology” to “Social” does not make it a science.

  100. Many of the comments here are cracking me up. How can anyone take this seriously — even before the dubious science???
    The authors of this “study” are taking a complicated exercise involving keeping or not keeping a kid’s favorite stickers — extending it to a variety of kids of a variety of ages from a variety of cultures — and saying it measures “altruism.” Without any control for culture, age, intellectual ability, maturity, temperament or anything else. How can anyone possibly think it measures ANYTHING? At the very least, the kids should all have been the same age or developmental age.

    This is then “correlated” with the “religiosity” of their homes as defined by how their parents answered questions about their own abstract experience and frequency of religious services. Is a Pentecostal church service equivalent in any way to a Muslim prayer service? Do all “religious” families have the same sort of home life? Do the children have religious instruction? If they are not similar in some specific way, then that’s equally useless.

    Put the whole thing together and you get… NOTHING.

  101. Briggs,

    As Feser would state, this just reeks of amateur philosophy.

    To be honest, the only only takeaway from this study confirms is that the researchers believe atheism is false?

    I guess they can believe what they want, but well you veer into morality, that’s a big no-no for atheism, regardless what you believe or not.

  102. It never ceases to amaze me just how far some will go to put a nasty label on being religious! Don’t they have anything better to do?

  103. Could you explain the religiosity scale to me? Is it really a – 1, 0, and a 1? I don’t think I’ve ever seen a graph so confusing.

  104. I don’t know anything about the SCIENCE of such sociological studies, but I had a good time demolishing this study based on first-hand observation of kids. As follows: “Seriously? A sticker exercise administered to a couple of dozen selected children from each of these countries, with vastly different cultural, social, educational, and governmental norms, is supposed to demonstrate valid conclusions about morality? The subject sample is so small it could not possibly maintain the most basic standards or controls of a scientific study to allow for significant variables such as differences in children’s living conditions, family circumstances, nutrition, I.Q., etc., even within a single country, not to mention across the world. And how does one derive comparisons about punishments for bad behavior from a sticker exercise? Did they have to fight each other off to acquire the available stickers? — in which case, did the study allow for valid comparisons between children from large families vs. those with few or no siblings? Were the children all exactly the same size and weight? Did some or all play sports? Team competition (soccer) or individual (track and field)? Were some musicians, careful of their hands? And what about that well-documented behavioral determinative, birth order? — aggressive oldest, put-upon youngest, or classic peace-making middle child? What were the controls to measure differences in behavior and attitude towards regulation and authority based on whether they lived in a police-state or a snitch culture?! Having a foot in both countries, I feel safe in saying that a sampling of two dozen kids from just Canada and the U.S. — culturally similar as they are — would collapse under scrutiny on the basis of relative populations alone. On its face this study is a joke — I would suggest that it was devised to demonstrate prior conclusions. It may have brought a smile of satisfaction to the faces of Beast readers, but that should be followed by a blush of embarrassment at having fallen for such snake-oil nonsense.”

  105. There were two measures of “altruism” in the study. The first was sharing stickers. The second was how severely the children judged depictions of interpersonal harm. Children from religious families were more likely to be more severe in their judgments. But how is the latter a depiction of “altruism” or “meanness”? Don’t we want our children to stand up to bullies? Is it altruistic to be unconcerned about the misbehavior of others?

    I would also note that there is no information presented about any comparison within individual countries. There may be differences in societal attitudes between the US and China, for instance, which are coming into play here.

  106. Praise the heavens there are still some intelligent people left in the world. I thoroughly enjoyed/appreciated your review, and your witty sarcasm made it that much more entertaining. Keep it up. You’re one hell of a smart dude.

  107. Daniel: The dictator game was originally used by Kahneman to “offer evidence against the rationally self-interested individual (sometimes called the homo economicus) concept of economic behavior” (Wikipedia). Let’s try this:

    “Religious children are more rational than their secular counterparts, study finds: Religious belief appears to have positive influence on children’s rational economic behavior…”

    Actually, Wiki continues to state that “precisely what to conclude from the evidence is controversial.” Some think it’s altruism, some think it is socially acceptable signalling to the researcher because
    greed is frowned upon.

    “Religious children are less driven to impress strangers than their secular counterparts, study finds: Religious belief appears to have negative influence on children’s need to impress strangers…”

    You give a child a windfall of stickers. You start talking about sharing those stickers. Is the child immediately triggered to see, in the mind’s eye, all the children it would like to share stickers with? The child shares them out evenly: Two for Daniel. Two for Janet. Two for Briggs. Two for researcher’s pick. Keep two for me. That makes eight in my envelope, two in researcher’s.

    “Religious children are have more friends than their secular counterparts, study finds…”

  108. I doubt very much there would be any measurable difference between religious and secular children on any replicatable study. Everyone seems to think, even children, that they are morally superior beings.

  109. I saw a study in which 5 year olds shared stamps when they were monitored, but not when they weren’t monitored. Didn’t we once have an ‘age of reason’ at about age 8?

    The discussion refers to data showing children from religious families being more ‘punitive’ but it was really only one religion, the most structured one.

    It may not be a negative for young children who are developing a conscience (internalizing) to over-react to wrongs.

    The study doesn’t attempt to explain why, if the religiously raised are comparatively less generous as they age, they become more charitable as adults.

  110. Briggs, thanks for your review. You raised every issue I had with this “study”

  111. Ye Olde Statistician

    November 10, 2015 at 7:51 am

    the point of science is to change, to always be questioned.

    Except for Darwinian natural selection, global warming, et al. Actually, scientists are just as susceptible to confirmation bias, group-think, grant-hunting and other such things. Experimenters after Millikan adjusted their results trying to match his, and only gradually did they creep toward the now-accepted value.

    the unit under test works; even the definition of “works” has to be clear

    I once had a two-hour discussion with a tooling engineer who wanted me to design an experiment to discover if his new design “worked.” LOL

    The marvelous thing about these experiments is the way that they call things “measurements” and “instruments,” as if they were doing something akin to obtaining the thickness of a film layer or the viscosity of a melt.

    There can be no science of any hard empirical variety when the very act of identifying one’s object of study is already an act of interpretation, contingent on a collection of purely arbitrary reductions, dubious categorizations, and biased observations. There can be no meaningful application of experimental method. … It is as if he imagines that by imitating the outward forms of scientific method, and by applying an assortment of superficially empirical theories to nonempirical realities, and by tirelessly gathering information, and by asserting the validity of his methods with an incantatory repetitiveness, and by invoking invisible agencies such as memes, and by fiercely believing in the efficacy of all that he is doing, he can summon forth actual hard clinical results, as from the treasure houses of the gods.
    — John Bentley Hart

    would have made Aristotle seem scientific, yet several of his deductions, several of his methods, and his hubris set back science and human advancement a couple of millenniums.

    Oh, the old Stagirite was scientific enough. No one else in history ever invented natural philosophy as a coherent discipline with a body of knowledge and methodology and no culture ever approached modern science save by jumping off from Aristotle. It didn’t set human “advancement” back, because there had been little or none previously. (And “advancement” thinking was a medieval thing anyway.) The main drawbacks were he sometimes had bad data and he was unable to measure most things. Yet he was one of the few ancients who insisted on empiricism. He held a conviction that an experiment, being artificial, would interfere with the natural course of things. But Darwin was awestruck at his works in biology, and even his physics is accurate as applied to motion in a plenum. (He overestimated the effect of air resistance on falling bodies.)

    Besides, most human advancement was due to engineering, not science.

  112. Even if you ignore Wm. Brigg’s methodological concerns (which I think are legitimate), there are still legitimate statistical concerns.

    The article reports a multiple linear regression analysis with first-order effects only and no regression diagnostics. Even those most favorable to regressing Likert-type data understand that it’s only tenable conditional on certain desirable psychometric properties of the scale like internal consistency reliability and a coherent factor structure (not reported) and conditional on proper model behavior (also not reported).

    That first-order effects only were reported ignores the possibility of higher-order interactions that would make those effects interpretatively meaningless.

    Since the criterion variable (number of stickers remitted) is a count variable, there is no reason to suspect it should be normally distributed: generalized linear modeling specifying a Poisson or negative binomial response distribution probably would’ve been more appropriate.

    If OLS multiple regression continued to be used, residual analysis should’ve been conducted and reported (or at least noted!) to ensure model validity. I’d wager lots of money that this study produced a very ugly residual plot, probably not even fan-shaped but consisting of diagonal slashes owing to poor psychometric properties of the response variable.

    At least one variable (country of origin) probably should’ve been treated as a random effect with participants nested within it; there is some evidence that accounting for the naturally hierarchical character of data can negate or even reverse apparent effects.

    That’s just what came to mind this morning before coffee.

  113. Briggs

    November 10, 2015 at 8:45 am


    Where did that JB Hart come from? It’s not David, by any chance?

    Update: Found it.

  114. Dori: It does matter if this study had a valid conclusion or not. There is no way tell. The procedures for science were not followed and therefore yield nothing scientific in the answer. Yes, it shed light on poor scientific procedure. This is NOT about whether the conclusion is valid–again, we DONOT KNOW. Which is exactly where we were WITHOUT the study. Therefore, it has no value other than being a bad example.

    Julie: There were two religions–Christianity and Muslim. There were more Muslims than Christian participants.

    All: The part about punishment is not addressed by Briggs. If you go to the original study, this is stated:
    “Moral Sensitivity Task
    In this computerized task, used previously with children in both behavioral and functional neuroimaging studies [19], a series of short dynamic visual scenarios depicting interpersonal harm (e.g., pushing, bumping) was presented.”

    If you read the entire study, this more accurately measures the belief in communism or progressivism. The children who did not share were labeled “bad” because they were not part of the team. Sharing or labelled as outsider. Technically, that’s not altruism, but it may have played into the results. The children who believed in punishment for wrongs were mean. You see that now in the progressives–it’s not your fault and anyone who says it is is just a Republican Conservative.

    A paper that discusses some of the problems with these methods:
    I am not going to discuss the points in the study, but it does discuss much of what is being discussed here.

  115. Sheri, there were Muslims and Christians in the study but only one group, Muslims, had a statistically harsher response to the level of punishment compared to the nonreligious.

    That means only one religion of 5 is in the second finding.

    Yet in the discussion part of the study it clearly says “A second major finding from these data is that religiosity affects children’s punitive behaviors.’

    Tsk. Tsk.

  116. Ye Olde Statistician

    November 10, 2015 at 12:03 pm

    @evecon12: Nicely and concisely put!

    @Briggs: Aargh! Your enemies have turned their attentions on me.

  117. Julie: The bar graph shows Christian and Islam children both exceeded the nonreligious children for judgements of meanness. The same is true for “sensitivity to injustice”. What am I missing? Both have p values <.001.

  118. There are much bigger elephants in the room for this study. Much, much bigger.

    1) The study was 100% urban. All the kids came from cities. 1 city per country, in fact. Not a broad, spread over the world study. 6 cities. That’s it.

    2) The authors’ didn’t compensate for economic factors. They did compensate for education level of the mother. How the hell that’s a good proxy for socioeconomic status spread over 6 countries, I have no idea.

  119. Sheri,
    Where it says ‘thee were no significant differences between children from Christian households and non-religious households.”

  120. Scott Adams, in his book “The Way of the Weasel” defines altruism as: a hallucination that weasels are capable of becoming non-weasels.

  121. Sheri,

    Where it says in the study, “there were no significant differences between children from Christian households and nonreligious” ones.

    This refers to the harshness of punishment scale.

  122. I’m also questioning the scientific method I the delivery of the tests. It’s always easy to say; since it’s a “scientific” study then the testers were following the scientific method 100%. However, we have here a psychological test, which ascertains a certain behavioural response of children based on “quantifiable” input.
    To be truly scientific about it they should have also introduced tests that verified that their claim wasn’t not correct. ie introducing variables that indicated no variance of results.
    Also, the testers would have been hell bent on getting a certain result and would have introduced this lean into tests. Which is why we have such things as double blind tests.

  123. The doctor and dentist offices that I take my children to always have stickers available, and the patients are not limited as to how many they can take. My kids who are 7 and under will often try to get several stickers, but they plan to share with siblings at home. My 8-9 year olds will probably only take one for themselves (if it was for a character they really like) and some for another sibling at home. From 10 up they could care less about the stickers. They would probably grab one or two for younger siblings and none for themselves. Also, kids have definite taste when it comes to stickers. You can have twenty different styles and maybe only one will appeal to a kid.
    So I’m wondering if the stickers were any good and also if any of the test subjects may have had younger siblings to whom they planned to bring home stickers. A five-year-old child (or older) in a faithful Catholic family may have a few little brothers or sisters that they will want to give stickers.

  124. Just to add to the reply count:

    1. Good article
    2. Much humour at troll replies
    3. Here in the UK child bothering is a crime

    That is all.

  125. I first saw this article mentioned in The Economist. In addition to the problems with definitions and statistics, I also thought that I don’t really give a rat’s patootie about whether the kids are altruistic (however defined). How they will deal with life as adults could at least be interesting, though the techniques used are even less adequate to answer THAT.

  126. Matt:
    I am sorry I have been away so long. I just happened on this. I don’t have time right at the moment to comment in detail but I agree with you on so many levels and points.
    The two most basic for me are that simply on its own terms the model is grossly under-specified and, epistemologically, the method fails to ask the children what led them to chose the number of stickers they gave away. There are many more issues here.
    On the first point, and as others have noted, the addition of another obvious variable could easily produce a completely different set of conclusions. One that was already mentioned is “Do you collect stickers?” (I think a binary response should be sufficient.)
    The epistemological point speaks for itself.
    I will be back. This is too much fun!!

  127. Quantification in contemporary psychology is a politically instituted moral imperative for psychologists to look “scientific”—rather than a rational choice (Porter, 1986, 1995). Mathematicians have pointed to the possibility that the use of real numbers to characterize psychological phenomena is unlikely to be useful (Rudolph, 2006; Zeeman, 1962). Yet once a moral imperative is in place, rational arguments fade and social cohesion processes guide a social science (Toomela & Valsiner, 2010, p. 328).

  128. Irregardless of your views on this article, I think we all could have done with out this sexist remark: “And most importantly, don’t throw a temper tantrum and (for you men) cry like a little girl”.

  129. One more note:

    Just as a man on a desert island was held to illuminate the moral order so a rat or a monkey or student pressing a bar was thought to illuminate the brain. . . . The belief that the truth can be laid bare by parsimonious means is inherently handsome. The conceit that this can be done by means that are trivial is perhaps inbred and even a little decadent, but attractive nonetheless. (Hudson. 1973. pp. 40-4 I).

  130. Despite your flame-war attempt at rebutting this argument, I am going to go ahead and do some study on the topic of altruism and growth, and I am going to have as a preset the education in critical thinking, ethics and values and philosophy that I already have. I will need include statistics, of course, lots of them, and I may come to the same conclusion as you. However, unlike you, the author, I will phrase my findings in such a way as to not start a flame/comment war.

  131. Julie: You’re right. I read past the statement “Moreover, children from religious households also differ in their ratings of deserved punishment for interpersonal harm (F(2, 847) = 5.80, p < 0.01, ?2 = 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." Apologies.

  132. Joy: Is it sexist if I say it? Just curious.

  133. I started reading this because I had seen the article in The Guardian.

    But I stopped when I got to the following snarky characterization: “Science means unquestionable truth.” No, it doesn’t, and you must know that. No scientist claims such. The petulant whining of the opening paragraphs is really tiresome.

    I hope the rest of article was good.

  134. Matt,
    The p-values may be wee, but what about the r-values which would seem to be quite consistent with the data scatter? Isn’t the relative explanatory power of a regression described by r^2 rather than r? Then, why would anyone publish a study wherein their regression fails to account for 83 to 97% of the variation in the data?

  135. Hi Briggs,

    Just wanted to thank you for taking the effort to break the argument down point by point and replying to your readers in a similar manner. I read in another study that such behaviour on the internet is considered pointless.


  136. Well, as soon as I got to the bit where it described how the test was done, I realised how meaningless the results were. If anyone can’t, then there’s something wrong with them.

    As a simple thought, the child KNOWS they can simply take all the stickers and share them out later, why should they make that decision here and now? Maybe they had 3 friends and wanted to share their stickers amongst all their friends rather than just have to pick one of their friends to give them to. Maybe they wanted to check which of their friends had also received stickers before deciding who to give stickers to. Maybe they wanted to give some to a friend who was not at school today because they were ill. Who knows? It’s a bizarre, and ill-conceived test.

  137. My (late) take on this is

    – “Harshness” was “measured” according to how negative the reaction to described situations of “interpersonal harm” and how strong the punishment demanded was. So… religious kids have a stronger sense of justice?

    – The definition of “Altruism” has already been justly ridiculed here.
    (Wait, about how the “game” was set, did they really expect 12 years old to buy the absurd explanation by researchers?)

    – Many hailed this as a study with a “large” sample (n=1170). Problem: in the analysis this is divided between (controlled over) different religion scores (or 3 main religious denominations), 6 countries, a very wide range of ages (5-12, why the hell did they not study a single age?) and the level of education of the mother (WTF, in the whole paper they call this the Socio-Economical Status of the family, until you read the fine print at the end! Why on Earth did they just not ask the SES?!). The sample doesn’t look this large anymore. Moreover, even in the Supplemental Information one cannot find *how* the sample was taken, which is no small deal. I definitely second Pat.

    – Beyond the pure lunacy of the so-called “SES proxy”, there’s an important variable they did not control: number of siblings.
    OBVIOUS ALTERNATIVE THEORY: religious people are well-established as having more kids, so kids from religious households are more prone to count in the stickers share themselves AND their brothers/sisters.

    – There’s the usual “measure the unmeasurable pointed by Briggs, with “religiosity” normalized in a numerical scale combining arbitrarily TWO DIFFERENT questionnaires (“frequency” vs “strength”) and only THEN arranged in a LINEAR analysis. This and the statistical concerns by evecon12.

    – The authors try to convince us of the goodness of religious predictivity of stickers-sharing by showing a plot (fg. 2, which, by the way, does not control on the other parameters, so it’s useless even within the study itself) which really needs no comment.

    – In the same vein, they show uncontrolled histograms (fg. 3-4), which again mean jack nothing.

    – Despite showing really small effects, the paper goes on ignoring the whole “correlation is not causation” maxim and constantly speaks in causal terms, without considering alternative interpretations (or any of the above limits).

    – In the end, a first name author who immediately goes to Forbes to comment stuff like “It’s not like you have to be highly religious to be a good person,” Decety points out. “Secularity – like having your laws and rules based on rational thinking, reason rather than holy books – is better for everybody.” can be hardly considered unbiased.

    Compare how poor this study published on Current Biology (WTF?! Where’s biology in this?!) was with its neutral, purely academic last line: “More generally, [these results] 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.”

  138. L: I wondered the same thing–which is why I thought this tested progressive behaviour rather than religiosity (unless we count progressivism as religion). The children did not advocate just punishing someone because you didn’t like what they did–there was a reason for the need for punishment. Actually, progressives advocate punishing those who don’t think like you do.

    It’s interesting the researchers did not contemplate the idea the kids were saving the stickers for siblings, yet many here did. Who’s the altruistic ones in this? Remind me again.

    There needs to be a study on people who refuse to read an article through but throw in with angry, nasty or indignant comments anyway. Funny they couldn’t bother to read the article, but had plenty of time for commenting on something they did not read. Some of the comments basically say these people are victims of “triggers” and their fragile psyches cannot take any disagreement or sarcasm. Said individuals should limit themselves to reading liberal papers so as to avoid that problem. Plus, if you are bothered by tone there, they might actually rush in and accommodate your discomfort. (Schools really need to teach the meaning of the word “sarcasm” because apparently many people have no idea what that is.)

  139. Ye Olde Statistician

    November 11, 2015 at 11:55 am

    why would anyone publish a study wherein their regression fails to account for 83 to 97% of the variation in the data?

    Because they’re not real scientists or engineers. Their data are so vague and ill-defined that they get all warm and fuzzy to find a wee p-value that says only the r does not equal zero!
    Eric: I stopped when I got to the following snarky characterization: “Science means unquestionable truth.” No, it doesn’t

    You have obviously never questioned Darwin’s account of evolution or the modern, model-based account of global warming. Scientists are like mothers. They get all angry and defensive when you say their baby is ugly. See Feyerabend for details. Or see L’s comments, above.

    Also, FYI, this: https://en.wikipedia.org/wiki/Sarcasm

  140. Very clever Eric. You almost had me. Imagine! A Grauniad reader offended by snark and petulant whining. How would they ever get through a single edition? But really, you must start using sarcasm tags.


  141. Nice post, bro.
    I have zero knowledge about statistics or how you should conduct a study like the one you’re talking about.
    You give us sources to support what you are saying. BUT, your source is your own blog. I would suggest to change your sources.

    “Why I am Right Update P-values are rotten evidence for anything (click here to learn), (2) Regression is deeply flawed and not what you think (click here, here, or here to learn), (3) Probability models do not prove cause (click here, (4) Asinine studies like this are common (click here) or here). ”

    only one link of the above part is not taking us to your site.

  142. L
    Nice extension of the discussion. A quick look indicates hat the first author, Jean Decety, has published in the same journal a number of times. Somebody has to have the first article in a journal but why would it be this article in a journal titled Current Biology. Am I being too cynical?

  143. The study’s correct conclusion is “children whose parents self-report going to church and believing in God are less likely to give away stickers to children who have none”.
    I think even that conclusion may be off. A more precise one might be,
    “children whose parents self-report going to church and believing in God are less likely to give away stickers to scientists who claimed they were to busy to visit other children and who claimed stickers returned to the overly-busy scientists might somehow be given to those other anonymous children in some unspecified way”.

    Some children may not have given the stickers back to the scientists because they didn’t trust the scientists to give them to other kids. For all we know, the kids figured if they kept the stickers, they could give them to unlucky kids themselves rather than leaving that job to the supposedly too-busy to play scientists. The fact is: we don’t know the kids motives for keeping of giving away stickers.

    It might be interesting to learn whether the stickers placed in the envelopes ever were given to other children in the school. They may have merely been counted and then tossed away. Call me a cynic, but if I were a betting gal, I’d bet the children who did not get to play were not given stickers from the “give away” pile.

  144. Lucia:
    As you probably realize, you raise an issue that is part ethical and part epistemological. I recently plowed my way through Dan Ariely’s The Honest Truth about Dishonesty, which is full of similarly designed experiments. The irony of an experimental design that requires dishonesty seemed to have escaped Ariely as much as it escaped Decety. The absence of any inquiry as to how the 5 year olds thought about the exercise compounds the irony even further.
    The late Herb Kelman famously raised these issues about the Milgram coercion experiments. A great deal of self-examination followed but psychological/social psychological experiments involving human subjects continued more or less unchanged. One has to wonder at how academic boards responsible of experiments with human subjects would have allowed Decety to use children as young as 5 and 6. The mind boggles.

  145. @Lucia
    It might be interesting to learn whether the stickers placed in the envelopes ever were given to other children in the school. They may have merely been counted and then tossed away.

    Lucia, If you are correct and I would bet that you are, then a follow-up paper entitled “The Negative Association Between Lying to Children About Distributing Stickers to Less Fortunate Children and Children’s Altruism Around the World” would be in order. In fact, such a study should be carried out across the child’s lifetime and regressed against all future potentially negative behaviours e.g. criminality, cruelty to animals, sour dispositions, bet welching, dyspepsia, cheering for the Cubs, etc. 😉

  146. I don’t understand why using the DRQ religiousness scale with regression is a bad idea. Can you explain this please?

  147. C ,
    even if one was to be ok with quantifying the unquantifiable (which begs the question, is a “religiosity” of 0.8 really the double of 0,4 and the opposite of a -0.8?), let’s take a look at the very paper on the DRQ, cited by the authors:

    The DUREL [what the authors call DRQ] measures each of these dimensions by a separate “subscale”, and correlations with health outcomes should be analyzed by subscale in separate models.

    What the authors do instead is defining “religiosity” as a normalized numerical scale combining arbitrarily TWO DIFFERENT questionnaires (“frequency” vs “strength”) and only THEN arranged in a LINEAR analysis.

    Goddammit, it makes me sick

  148. Another thought: the researchers made up a rather contrived story about why they had to ask to the children to SECRETLY share some stickers, and then promptly went to count how many any specific kid had shared. It is well attested that kids, like people, share more if they know they are watched: couldn’t it be that some suspected what sort of lying f*cks the researches were, further confounding the results?

  149. Brignell argues that sarcasm tags shouldn’t be necessary. My apologies to Dr Brignell, chains like this make me think they are all the more important. Not only do we need sarcasm tags, we need escalated sarcasm tags. We also need “Yes I am seriously taking this side” tags. Then there needs to be the moderators of the tags to assess whether people are using the tags to troll with.

    Epidemiology doesn’t work very well. They are still trying to isolate the Chipolte e-coli initiation point. I want the Chipolte opening up in my area to actually open up. I haven’t found out why they are the bees knees yet. Epidemiology is failing yet again at finding the source. The “usual” suspects didn’t pan out.

    I am biased when I read Briggs work. He always seems to be pointing the same way I am pointing. I haven’t quite figured out if I am just pointing the the same way he is pointing. I do know that I am not pointing the same way that pretty much an study that has a p-value in it is pointing. I laugh at wee p-values because people get excited about them. I just keep looking at the data that contradicts the p-value (or the Big RR, OR, HR) and wonder at how they can ignore the uncertainty.

  150. bernie1815: Perhaps they did have parental permission for the “game”. Lewendoski did the same with his global warming “studies”, including minors in the data. Even after complaints, the data does not appear to have been removed and I suspect he could do it again and if no one closely read the data, he’d get away with it. It’s a serious problem–scientists can no longer be trusted.

    I agree with Steve E–a follow-up study should have been done here to see what the kids actually did with the stickers. That would have been more interesting and telling.

  151. I took psychology 101,102,103, 201 in college. I got extra credit if I participated in the psychological studies. I didn’t really care what the results were. The one thing that all the studies had in common is that the statement of what they were studying provided to us was always a lie. I understood the necessity of the lie. It has always bothered me though. The lie always twisted the results a little. Unraveling the twist never quite seemed possible.

    The dance never ends.

  152. brad.tittle: I did studies as part of consumer psych in college. It is possible to set up experiments that tell the participant what is being tested–especially when comparing products such as taste of brand name versus generic. It’s tougher when looking at behaviours, since observation changes behaviour, so there are some “lies”. Those can be held to a minimum by not telling the participant what the study is about or what is being studied. A double-blind experiment helps. It could be considered a lie of omission, I guess, but I never really thought of it that way.

  153. @Sheri
    The whole experimental design for social psychology is a** backwards. If we break this one down, we have to ask why are they talking to children as opposed to other “religious” or “non-religious” populations. You presumably could get to the same findings simply by asking a random sample of people to show their tax returns and talk through their charitable donations. You don’t have to use subterfuge or lies. Simply ask individuals to explain their donations.

    As for Lewandowsky, please do not get me started.

  154. bernie1815: I agree this study is totally messed up. It did not have to be that way. There are much better ways to measure generosity and religiosity, assuming we can define them in an acceptable way. (I’m not sure that telling the kids this was a “game” is a lie—it was a game. What the kids were not told was that it was also part of an experiment. One really cannot tell the kids this or it skews the results. What do you consider a lie here?)

    The study I linked to here covers some of the problems of self-reporting on generosity:

  155. @Sheri:
    This can take us far afield. For me, whenever the experimenter does not disclose the purpose of the experiment in ways that the participant can understand then he or she is deceiving (i.e., lying to) the participant. The earlier issue of where the stickers went after the experiment strikes me as potentially an explicit lie but it is part and parcel of the larger deception. It also strikes me that the experiment and its results are far more potent when the participants understand its purpose and can essentially knowingly cooperate with the researcher. Imagine the potency of the Milgram experiment if the participants were in fact informed as to the purpose of the experiment – just like the concentration camp guards were.
    As I mentioned, Herb Kelman triggered a big debate on these issues when he challenged the ethical standing of the Milgram experiments. (I took courses with Herb as a graduate student in 1974-5.) I also worked with Chris Argyris. His book “Inner Contradictions of Rigorous Research” has played a huge role in how I view these types of experiments. Chris would view the current study as having limited explanatory or practical value. He would likely argue that Decety would have done better to ask questions such as how do individuals decide how many stickers to share, what are the differences among those who share few stickers and those who share a lot and what do the experimental participants say would increase their willingness to share. Such a reframing of the research does not exclude the possibility of differences by religiosity. It offers an opportunity, in fact, to see the extent to which religious affiliations get manifested in both the actions and thought processes of research participants. It also makes it almost impossible to treat the participants as pawns (subjects) to be manipulated by the experimenter. It forces the experimenter to come to terms with the fact that human participants approach such experiments with theories of what the event (experiment) is about and to address any “demand characteristics” that might be perceived and that might impact individual decisions and behaviors.

  156. bernie1815: I understand your objections to some research. Your examples are of bad experiments. There are ways around that—without explicitly lying. However, there are times when one cannot tell the participants the reason for the study. Humans tend to want to please and if they know what you are looking for, they tend to deliver just that, or deliberately avoid delivering what you want. Again, there are ways to avoid this. Experiments involving people in rooms doing things that are not part of life are not a good design. As noted, we could go way off the subject here, though experimental design is still on topic, I think. Perhaps I can find a study done properly and submit it to Briggs later on. For now, I’m out of time so I have to sign off. This is a most interesting discussion and I’ve enjoyed the exchanges.

  157. @Sheri:
    Likewise. 😀

  158. Yeah, you guys keep on battling each other. All the intelligence and facts in the world matter not one bit when you all starting arguing like adolescents.
    Hard to take anyone seriously then. Everyone has an opinion. This study or critique of it cannot be taken seriously in the first place.
    Carry on…

  159. Many interesting observations in the post and comments. A further question which occurs to me is whether there might be a correlation between country and religiosity which happens to parallel a correlation between country and choice to hold on to stickers. i.e. if there is cause, might it be more to do with country than religion?

    I also wondered whether the scatter plot might hold some structure other than the obviously inadequate planar model used by the authors. One commenter above made suggestions along these lines – but how does one see such structure in a single 2D projection?

    A further question concerns correlations to individual “subscales” of the DUREL questionnaire on religiosity. The Duke authors of this questionnaire explicitly recommend this in any analysis, observing, for example, that subscale 1 and subscale 2 could cancel each other out if one attempted to consolidate them into a single-number metric.

    I have e-mailed the first author of the study to ask whether he might share the underlying data. This may be deeply naive, but it will be interesting if he does – then the questions above might be tested.

    (Anyone thinking that this would be pointless given the meaningless of the data in the first place is right. But the bit is between my teeth.)

  160. Tom:
    You can ask, but I doubt that he will release this type of database any time soon. It offers too many opportunities for additional papers by one or more of the current authors. An alternate strategy is to ask for other articles and research reports written by any of the authors that are based on the same set of data. This might give a better view of what else is in the dataset and raise questions as to why they were not included in this analysis, i.e., the under-specified model issue.

  161. I stopped reading this when it stated “human race”. Humanity is NOT a race, nor are their any races of humanity, we are a species. How hard can it be…

  162. Ye Olde Statistician

    November 16, 2015 at 8:58 am

    There goes Seamus MacManus’ “The Story of the Irish Race.”

    Unless a race is a group showing common features within a species, pretty much the same as “breed” or “strain.”

    Politically, of course, no such thing can be allowed, since knuckleheads have often tried to tie essential characteristics to them, such as parsimony to the Scots or stolidness to the Germans.

  163. Again someone says indignantly “I stopped reading this…” and jumps right in to comment on why they did not bother to read the article. Is this a short cut to commenting or what? At least read the whole article and then enumerate your objections.
    ( Maybe we do need trigger warnings so people don’t start reading and then waste the two minutes it took before they quit and commented. Perhaps a quiz before people can comment so everyone actually reads the article?)

  164. > I stopped reading this when it stated “human race”. Humanity is NOT a race, nor are their any races of humanity, we are a species. How hard can it be… <

    This kind of pseudo-precision makes it hard to have a reasonable conversation. But, of course, if you stopped reading, you're apparently not interested in a reasonable conversation. "Human race" is a phrase with a long and distinguished pedigree. (A primary definition of "race" in the OED is "a group of persons … connected by common descent or origin." What's your definition of a race?) Have you expunged "sunrise" from your vocabulary because it's anti-scientific?

  165. Neophyte question: the paper doesn’t define its terms (which, I understand is normal in this kind of paper). We are just supposed to know what is meant by “r” and “Msharing” and “?standardized”, etc.

    So I am reading up to figure out what I am supposed to think these terms mean. I think I understand that when a paper speaks about “r”, then it is talking about a correlation coefficient. So saying that r=-0.173 for the relationship between parent religiosity and sticker count means than there is not a particularly strong correlation. Am I reading this correctly?

  166. r is the correlation coefficient. The closer to one or negative one the value is, the stronger the correlation. (r ranges from -1 to 1). So -.173 is not a strong correlation.

    I tried scanning through Briggs’ classic posts for standardized, but didn’t see anything right off hand. This is from a UCLA page on statistics:
    “A standardized variable (sometimes called a z-score or a standard score) is a variable that has been rescaled to have a mean of zero and a standard deviation of one. For a standardized variable, each case’s value on the standardized variable indicates its difference from the mean of the original variable in number of standard deviations (of the original variable). For example, a value of 0.5 indicates that the value for that case is half a standard deviation above the mean, while a value of -2 indicates that a case has a value two standard deviations lower than the mean.”

    M(subscript)sharing is the mean value for sharing. It’s followed by the SD, or standard deviation.

  167. I have an alternate interpretation of the results from the Dictator Game: “religious” kids are less likely to be manipulated by lies told to them.”

    After all, I’d be willing to bet that the perfectly-adjusted-moral-compass “researchers” didn’t actually give any of those “taxed” stickers to children who did not participate in the study. Further, I’d also bet that the of-course-we-don’t-have-an-axe-to-grind “researchers” knew which of their subjects were “religious” as they “leaned on them” to be generous.

  168. Seentahna Bushyhead

    November 21, 2015 at 10:57 pm

    Here’s another reason why the study is flawed. The kids weren’t asked to give their stickers to an actual living breathing child. They were asked to give it to an envelope. How many 5 or 6 yr olds would understand that? If they were asked to share it with another child standing right in front of him, I’m guessing the outcome would be much different.

  169. I saw the study and scoffed, but good to see it disassembled by a statistical marksman. Briggs, you truly are the right’s official statistician.

  170. This authors attitude seems to support the article he is mocking. It is not very altruistic to call someone else’s hard work ‘crap’ so dismissively. Perhaps if he shared his points with (what seems like) more of an open mind I’d be more inclined to have more faith in his opinion

  171. People who complain that operationalizing such a thing as ‘altruism’, and that the experimental setup is not ‘all-encompassing’ and isn’t perfect are the kinds of people who just scream scientific illiteracy. Altruism has various interpretations and you can either try recording data on it or just say vapid qualitative things about it for another 5,000 years as our species has been doing so far. Let science progress. It’s not be-all-end-all. It’s just a stepping stone which can be questioned, replicated and falsified. So:

    OF COURSE altruism is not sticker sharing.
    OF COURSE 1,200 kids do not represent every human being in the world.

    But good lord can you just consider the statistical difference between religiously-raised kids and secular kids? It’s a great study and deserves publication, coverage by journalists and thought. My two cents!

  172. do1don:
    It is not that the operationalization of altruism here is not perfect – rather that it is stupid and from what I can see completely unvalidated. What are the odds that this study would have been (a) completed and (b) published, if it was found that children of parents who voice religious concerns were more altruistic than children of parents who did not voice religious concerns.

  173. Disciple Dave

    July 21, 2016 at 8:42 am

    Dear Briggs,
    Your ‘review’ is flawed on many levels.
    1: You are criticising an article on its scientific value. Yet; you define science as: unquestionable truth. (click to learn the definition of science)(1). You are extremely wrong here.
    2: Your entire review tells us nothing about the other methods that have also been used to quantify the participants level of altruism (besides the computer that has been used). Such as the Moral Sensitivity Task’s and The Questionnaire of Cognitive and Affective Empathy. One of the main contributors of this article is a respected Dr. who has been involved in this field of research with children for over 30 years . (click to learn)(2). I know this does not necessarily means that his experimental research is perfect. But if i were you, i would be cautious about disregarding their data mining as “crap” or claming that “altruism was not measured”.
    3: You fail to explain why it would be necessary to define “every possible nuance of the scientific concept of altruism”? I dont know why “no-greater-love” and “supreme sacrifice” are relevant for children between the age of 5-12. You introduced this factor, i dont know why and i think you are wrong to do so. A certain definition would also be extremely difficult to measure and compare. When we define altruism as ‘when you are predominantly concerned with the good of others’ I think the dictator game, MST’s and QCAE can be very well applied to quantify a certain level of altruism. You classify the data gathered as “crap” without investigating the methods properly it appears to me that you shout without knowledge. U were not there when the tests were conducted, a Dr. with 30 years of child psychology however, was. And i do believe he measured a certain quantifiable level of altruism. (It is not wrong to question the methods, it is wrong however to make these bold statements without knowing much about them.)
    4: I think the hypotheses of the main atricle is about investigating whether religously raised kids are more altruistic than kids who are raised as an atheist. Why is it then wrong to question the care-takers?
    I dont know whether the method used to compare the values is applied correctly, if you claim this is not the case i think we can trust you on that, its your field of expertise after all. I think their conclusion is fascinating because we would expect religious kids to be more altruistic, instead, atleast for these 1200 kids the atheist ones are showing a lot more altruistic characteristics.

  174. Richard Smiley

    July 27, 2016 at 10:55 am

    Let’s be honest: we like the original article because many of us have personal experiences from our own lives of some religious thickie looking us in the eye and telling us something on the lines of “I am going to heaven and YOU are going to hell!” WTF?! And then (in the case of the adult) seeing them arrested on a morals charge over their activities with their 14 year old daughter and two foster daughters. Schadenfreude indeed.

    Its something we want to believe because we’ve been provoked. Repeatedly. Religion is one of those things that some believers like to push down the throats of the rest of us. As you pointed out in your comments, there are whole fields similarly broken. Psychology comes to mind… 95% of all testing is designed to see if the test agrees with other tests, the arguments among practitioners about whether it is a clinical art or a science… ever done a rorschach inkblot test? Justice denied by some lawyer and psychologist blowing smoke in a criminal trial….

    Why are you so upset about this article in particular? Can’t you just look up beatifically and say “forgive them, father, for they know not what they do…?” or in your heart of hearts do you feel that religion something that is supposed to make you a better person… but doesn’t?

  175. This article is so very very angry in tone and completely derogatory to the study methods. Is that due to the author being religious?

  176. I did find one thing about this long comment trail very interesting.
    Seems Briggs spends an inordinate amount of time tearing apart anyone that disagrees with him, but gives a pass to illogical statements made by those that agree with him.
    I will give him one example to avoid being dismissed for not providing proof, and leave him to unearth the rest himself.
    A number of people pointed out that testing altruism among twelve year olds was not valid, as they don’t care for stickers. But that would not completely invalidate the test within that age group, although that group might seem more altruistic than younger children in the study because they would be willing to give away more of their stickers.
    The point is that Briggs’ lack of desire to correct and criticize those that agree with him gives hints about his narcissistic personality.

  177. I didn’t even try to get into your whole critique of the study since your view of science is deeply flawed. Science is not “unquestionable truth”. Science is always questionable. Scientists consider it truth until a better answer is found and proven scientifically. Evolution, gravity, electricity and even the big bang are all scientific “theories” and will be considered “truth” until a better answer is proven, tested and proven again.

  178. I like the way that at the end of your article you turn the tables and insist that anyone who disagrees with your assessment must ‘show exactly’ how you are wrong.

    That would be fairer if you had done the same in your review with respect to the study, but you didn’t. Instead you challenged the study’s assumptions without offering an alternative and then, without any expertise or context, went on to list a couple of articles that outline some limitations of statistical sampling.

    Your whole article smacks of defensive lashing out at any form of criticism. I suggest you may need a more considered and intellectually rigourous approach if you want to argue with a peer-reviewed study in future.

Leave a Reply

Your email address will not be published.


© 2016 William M. Briggs

Theme by Anders NorenUp ↑