z47\doc\web\2001\01\2sig.txt
Steven Tripp notes: 


In education there is something known as "Bloomıs 2 sigma problem";
holding time constant, students who are tutored score 2 sigmas above
students in (30 student) group instruction.


I have a colleague who anecdotally confirmed this by taking his daughter out
of school to "major" in gymnastics.  He and his wife (both former school
teachers) were able to cover a week's work in one or two days.


Assuming this is true and assuming you throw enough money at the problem to
achieve near-tutorial quality instruction, you might get a one-sigma
improvement.


Two caveats:  (1) I doubt anyone is thinking in these terms and (2) I sent
my son to an expensive prep school with small classes, but I doubt they

achieved even a one-sigma advantage over public school instruction.  Part of
the problem is that it's hard to recruit good teachers in math/science and
some other areas without MUCH higher salaries and prep schools don't pay
high salaries.  I suspect his math teacher didn't really understand the math
she was teaching.