The statistics I presented here are a little bit less processed than what you're describing. They compare only the full-time wage of women and the full-time wage of men, and they don't control for differences in occupation or differences in other background characteristics.
There is research that does that kind of comparison. To the best of my memory, when you control for all of those factors, the wage gap becomes 92¢ on the dollar rather than the 88¢ that we have here. So some of the wage gap can be explained by what we would call “observables”, but there remains a part that's unexplained.