I'm not sure about Judge Sotomayor's future as a quantitative social scientist:
While recognizing the potential effect of individual experiences on perception, Judge Cedarbaum nevertheless believes that judges must transcend their personal sympathies and prejudices and aspire to achieve a greater degree of fairness and integrity based on the reason of law. Although I agree with and attempt to work toward Judge Cedarbaum's aspiration, I wonder whether achieving that goal is possible in all or even in most cases.
That seems like a claim that differences in gender and ethnic background explain at least half the variance in decision-making, in other words to an r-sqaured of greater than .5. But she rests it on a mere finding of statistical significance in a minority of cases:
The Judicature Journal has at least two excellent studies on how women on the courts of appeal and state supreme courts have tended to vote more often than their male counterpart to uphold women's claims in sex discrimination cases and criminal defendants' claims in search and seizure cases.