That is interesting data.
I wonder, though, if the reason for not including additional data (represented in the table as "-") is... insufficient numbers?
The reason that I mention that is that I know that those numbers are not zero. I know this because DD is in one of those empty boxes. Er-- or 'was' in the 2009-2010 school year, anyway, which would have been her 8th grade year.
She was ten-- so does this mean that there were less than a thousand students like her in that census year?
Ahh-- nevermind. I found the footnote. That symbol means that the value rounds to zero (in thousands, presumably). So there were fewer than 501 students in any of those categories.
Particularly fascinating is 15 yo cohort-- there were significant numbers of those students at all collegiate levels, apparently, because there were enough of them to produce a "1" for college juniors, and a "12" (again, both in thousands, or rounding to that value in thousands, more properly) in first-year post-graduate programs.
I find that fascinating, particularly since the earlier values would not seem to support those kinds of numbers. Perhaps those PG students were never enrolled in regular schools when they were younger?
That makes some sense to me, given the barriers that we've encountered just in practical terms in having a child who has been accelerated out of the two-sigma distribution roughly apparent in the numbers for K through 12. If we were to have opted for an additional acceleration (which academically was a possibility), it would have stretched the fit to the breaking point in a couple of other areas. So at that point, homeschool or other alternative schooling begins to look like the most realistic and reasonable solution.
It would be fascinating to know, though, where those 12K kids in first-year graduate studies actually came from.
