The first (nature) paper appears to show that in market traders increased levels of testosterone result in more risk taking.

The third is mostly about fish and other animals, and doesn’t seem hugely relevant (though it’s interesting).

The second seems like the most interesting (MIT press article). They look at risk taking in adolescents. The paper seems to show that testosterone is linked to risk taking in males, but increased (natural) testosterone in females didn’t seem to show much increase in risk taking.

Overall, boys had very slightly riskier behavior than females. From the graphs the error bars are pretty big and the difference in risky behavior is quite small.

The difference doesn’t seem big enough that it would result in skewed gender ratios in jobs for example.

So, if this is well established in the literature, I’d expect to see better results than this. Are there papers that more really show that males are more prone to risky behavior and that this has a biological, rather than social basis?