These are two fascinating studies.
He found that students who attended high school with high-achieving peers performed better at graduation, passing more of their final exams than students who went to the same school in a different year, when the crop of classmates was weaker. Intriguingly, the effect of high-achieving peers was much more positive for girls than for boys. Jackson's results suggest that boys may, in fact, pass fewer exams when surrounded by high-achievers, while girls' graduation exam pass rates are helped by having bookish classmates.
Essentially, it is very good for girls to be surrounded by smart classmates. It is good for boys, too, as long as their classmates aren't too much smarter than they are. Being surrounded by poor performers has a negative effect on both genders.
Studies like these are crucial for education and a way to guide educators in how to make their programs more effective. I read stuff like this and think about the possibility of separating boys and girls into different classrooms. Girls can be thrown together for the most part regardless of level. Boys, however, should be grouped with other boys on a close scale of ability so that no one of them is that much better than the others cutting down the gap between the "best" and the "worst."
Unfortunately the trends seems to be more toward teaching for tests in reading and math and not worrying about anything else. I'm curious to see if Race to the Top will produce anything good, but I'm skeptical that it actually will. For as much money as we spend on education and for as important as it is, you would think that there would be more willingness for bold experimentation especially as we see our country falling behind other developing nations in terms of student achievement.