One of the reasons I went there—and I didn't mean to put you on the spot—is because it's very easy, and we heard this from some of the other witnesses, to look at your own programming and to say it doesn't have a gender component; it doesn't have any implications; it's not necessary. I understand there's a check box that you can check that says you looked at it and it doesn't apply here.
The training component is so vitally important because often on the surface, if somebody isn't very well versed in gender-based analysis—and this wouldn't be any implication here—you might say this doesn't have any kind of impact. Yet, when you delve into the actual results of some of the policy decisions, you realize that in fact there probably is a differential.
This is now to the departments, how would you go about looking at these hidden gender biases, where at the outset you would just think they don't apply?
Ms. Kennedy or Ms. Mitchell.