(Guest Post by Matthew Ladner)
Last week I showed you the above chart on the math and reading trends for students with disabilities. As you can see, a plurality of states land in the quadrant showing declines in scores on both math and reading. Arizona landed as one of the few bright spots, and when I ran Arizona trends on all six NAEP exams since 2009 (when NAEP reconfigured the Science exams, the national economy collapsed and shortly into the then new Obama administration):
In both the AZ and the USA cases, these are the NAEP trends of students attending public schools. Just think of how much higher those Arizona gains might have been if these poor families had not been distracted by the opportunity to attend a private school or to customize the education of their child through the ESA program…
Jay P. Greene's Blog 
How is “disability” defined? Hard to believe “comparisons” are valid.
“Student classified by school as having a disability (SD, IEP, or 504 Plan)”
Thanks. The classification processes are quite variable. It would be a mistake to read much into state by state or district by district comparisons.
Are you aware of anything in Arizona’s implementation of IDEA or NAEP that would lead to the appearance of unusually large gains? The implementation of IDEA does of course vary across jurisdictions, but random errors tend to cancel themselves out in large populations. It is especially good to keep an eye on inclusion rates in NAEP but AZ has been complaint, while that is not always the case in other states, which would tend to disadvantage AZ.
Matt, I have no systematic information that would be helpful. My general thinking is that practices vary within districts and between states. My anecdotal experience in AZ shows great variability but only between two districts and only involving two students. From WI days I can say with confidence that Milwaukee Public Schools applied classification methods that were quite different from those used in Milwaukee charters and private choice schools.