Here is a short critique of the recent religion and altruism study which has raised such a ruckus on social media. I do not agree with all of the points of the critique. The first point seems the strongest--the study did not control for enough possible confounds.
The second and third points of the critique seem weaker. The second is that the dictator game used in the study is not an effective measure of altruism. It's true that altruism is a difficult thing to measure, but the dictator game is widely used in research as a measure for altruism, and it's not obviously terrible (yeah, that's pretty faint praise--but this is social psychology, and everyone can agree that social psychology is tricky and messy and that we shouldn't expect transcendent perfection immediately in its methods and results, RIGHT?). For example, it has been used to show differences in sharing with strangers among people who live in different countries.
The third critique is that the effect size is small. But this is actually very typical for social science research. There typically are a lot of factors that go into causing complex psychological phenomena like altruistic behavior. It would be more of a surprise if the effect size of having atheist parents were large.
Still, even though this critique is not that strong, and even though I am an atheist and would no doubt enjoy seeing the study's findings confirmed, I am still skeptical about the way the study is being reported.
I am also disappointed about the way the study is being lionized by its defenders and demonized by its critics. It would be better if both sides could just acknowledge the limitations of the study but also acknowledge that it does provide some interesting or useful information that could help guide future research.
But no, people aren't like that. We primates have to band together with our in-group and verbally assault the out-group. The atheists instantly celebrate the study, the religionists instantly look for flaws or limitations in the study. Motivated reasoning, confirmation bias, the works. I.e., the usual disappointment.