https://uploads.disquscdn.com/images/5d8d7be525cd65b610fc22eb680e56c257afdf58a3b4b695081046daf0505257.jpg

https://www.sciencedirect.com/science/article/pii/S0960982216306704

It's a little discouraging that a flawed study like Decety's made it into a scientific journal.

"That is, 97% of the variation in altruism has to due with whether households are religious."

I think you left out the word "nothing"; "That is, 97% of the variation in altruism has NOTHING to do with whether households are religious."

From my point of view, the altruism study (funded by a pro-religion foundation) is utterly trivial and of no threat to parents who want to raise their children in a religious environment. I really doubt that you will see it seized upon by atheist or anti-religious organizations to somehow get parents to give up their "religiosity" for the sake of their chidren's "altruism." Even though the news accounts I read seemed to accept the findings of the study uncritically, I rather suspect that the average reader reacted to the study as described in the news media with much the same contempt as Dr. Briggs or Jane the Actuary.

Does the following at all change your opinion:

Robert Woodberry: “Religious Children are Meaner than Their Secular Counterparts” proclaimed a headline in the Guardian. “Religious Kids are Jerks” raved the Daily Beast. Hundreds of other newspapers and blogs touted similar articles: the Economist, Forbes, Good Housekeeping, the LA Times, The Independent. All these articles were based on a 4 ½ page research note in Current Biology by University of Chicago professor Jean Decety and six other scholars.

? I will note that you carefully qualified which negative effects would not be endured from this study being popularized in the way it was, with:

(1) an idiosyncratic, operationalized definition of 'altruism' being taken as equivalent to the popular conception of 'altruism'

(2) a confusion between 'religious parent' and 'religious child' (the study measured the former, not the latter)

Perhaps it would be wise to relax "which damage", and merely ask if "significant damage" will be incurred. I would like you to record your thoughts for the future, to make a prediction which, if falsified, will actually cause a change in your belief structure, in the belief structure which caused you to form the "it's not a big deal[, damage-wise]" opinion you seem to hold.

Sorry, but we, like Briggs, actually like sarcasm in front of such absurd "measures".

And l like my sarcasm to be funny and insightful. I also like my statistical analysis to have more depth. But that is just me. To each his own.

By your same logic, you did not explain why Risk! is not a good measure of propensity to violence, you only affirmed that.

We are having a conversation, and I was outlining points of agreement. Briggs wrote an OP. OP's tend to contain argument.

At least, by translating Briggs's words in "not sarcasm" you can make some cogent points (altruism is difficult to define, contain in itself aspects of "sacrifice, kindness, patience, love", is "nuanced"... all things absent in the 0-10 stickers-sharing definition.

So, I have to infer substantive criticism out of Briggs's articles?

Briggs says:

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

I suppose this is where I am supposed to infer everything that you just said. Actually. I was more focused on how Briggs makes a rather egregious error. We don't need to measure every nuance of altruism, but rather find a clever way to measure a significant portion of what it means to be altruistic or things that are related to altruism and more easily quantified.

For instance, we could measure what types of person donates more of their time to charity. Or donates more money.

Sorry, but we, like Briggs, actually like sarcasm in front of such absurd "measures".

By your same logic, you did not explain why Risk! is not a good measure of propensity to violence, you only affirmed that.

At least, by translating Briggs's words in "not sarcasm" you can make some cogent points (altruism is difficult to define, contain in itself aspects of "sacrifice, kindness, patience, love", is "nuanced"... all things absent in the 0-10 stickers-sharing definition. They combined different pseudo-quantifiers. The regression sucks and predicts nothing. They confound cause and correlation)

Then you just did not read the article

I've sadly read it two or three times. I even read the links Briggs inserted. I intend not to read Briggs in the future, but I read this article.

He doesn't explain every detail: as he says in his own blog, the paper is just wrong from head to tail, so he'd have to write much more than this.

Asserting something is completely wrong without telling us why it is wrong is useless.

But he points out the most obvious, irreparable issues: that the study does not measure what it claims to measure and that one of the variables in the regression has no scientific standing whatsoever!

All the bold and italics for emphasis really aren't necessary. Good, so it should be easy for you to quote Brigg's criticism with regard to this point.

These two flaws are not even statistical per se, but alone they make the study irredeemable: no matter what statistical analysis you come up, the wee p-values you calculate just don't mean anything.

I've been saying the flaws in the study are not statistical for some time. I don't see Briggs identifying them.

concluding that those who read comics are more violent: what would you make of the fact that whatever p-value I calculate is "statistically significant"?

No, because attacks in RISK are not a good measure of violent behavior. Now show me where Briggs does the same analysis of Dictator Games as a measure of altruism.

Again, it is easy for you to show me wrong. You just have to quote Briggs making a meaningful criticism undrenched in sarcasm.

From what Briggs has said I see not reason to believe that they chose bad statistical methods.

Then you just did not read the article. He doesn't explain every detail: as he says in his own blog, the paper is just wrong from head to tail, so he'd have to write much more than this. But he points out the most obvious, irreparable issues: that the study does not measure what it claims to measure and that one of the variables in the regression has no scientific standing whatsoever! These two flaws are not even statistical per se, but alone they make the study irredeemable: no matter what statistical analysis you come up, the wee p-values you calculate just don't mean anything.

Say I make a study that by a simple regression compares the number of attacks attempted at Risk! with the result of a questionnaire that puts together in a number the propensity to read comics, their grade of violence, nudity etc, concluding that those who read comics are more violent: what would you make of the fact that whatever p-value I calculate is "statistically significant"? This is exactly what's happening here.