[jkhdigital]

In general I agree with you, but the more I consider topics like this the more I see it as a problem of correlation strength vs. effect size. That is, at the population level there may very well be a strong correlation between birthday and certain personality traits, but the effect size is likely to be infinitesimal compared to everything else we know about what shapes a person’s psychology.

I actually think this is one of the ways that published science tends to distort rather than clarify our view of reality, as the strength of correlation tends to be overly emphasized as measure of quality or relevance.