That odd association seems pretty well-established by prior work. But the study at hand does seem like very well-done research to me: exploiting existing data t

That odd association seems pretty well-established by prior work. But the study at hand does seem like very well-done research to me: exploiting existing data to achieve a virtually perfect snapshot of a well-defined population at the scale of an entire nation. It seems like it neatly addresses all sorts of questions of sampling biases, sidesteps questions of statistical power, and relies on measures that are independently assured to be high-quality and comprehensive. With the added benefit of validating an effect in a wholly different subset of the population than the one where it had previously been measured: prior studies seem to have used mostly convenience samples of mostly men (mostly sourced from explicitly entrepreneurship-flavored environments).