Thanks for providing the data; really interesting stuff. I didn't read in detail, but I did scan through it. I thought several points were worth raising.

>>[cat's cradle] I mean no disrespect, but the issue is the ENTIRE ATTENDANCE area from which private schools can draw—not the school size.

The problem as I see it is the practical one of defining what that theoretical attendance area might be. One approach might be to find whether the farthest flung student commutes from -- a more accurate one might take the mean of all areas in a 360 degree analysis (not not assuming a simple circle/radius because of differing population densities.)

As the author notes, there is a lot of current unknowns used in trying to understand geographic data -- and there are even more issues when taking into account competing schools (e.g., Academic Magnet, for example, which was recently ranked as a top US high school by Newsweek) that draw from the same potential population of that area.

>>The reason we have A, AA, AAA, AAA is to correct for the draw WITHIN a school.<<

Is it to correct for draw or is it to correct for enrollment?

>>Again, there is research that private schools have advantages due to the ATTENDANCE area discrepancy, along with the inherent affluence advantage found in most private settings.<<

With respect to the affluence factor, I absolutely get it. Of course, as you've shown often with the poverty index, there is a wide distribution of affluence within various schools, both public and private, in our state.

After scanning through this, I find that I still don't understand the basis of the quantitative analysis for attendance pools using a theoretical draw model -- there just appears to be too many unknown variables.

Trying to compensate for theoretical draw would seem to me to be less effective than trying to compensate for actual affluence level of the enrolled population. That way we don't have very poor areas competing with very rich areas. It's my guess, and only my guess, that this would be a more effective way of equalizing teams than through some factor based on draw area.

>>The research, which I have posted before:

http://ohiosportsgeography.tadherold.com/hsfoot/publicprivatesxyears/publicprivate.htm <<