Why does this need to be “explained away” in the first place?  

“Convincing someone to revisit their own thinking is much harder if you only use *YOUR* own thinking to help them to do so. It is simply too easy for people to dismiss others’ thinking. ”

No?

I disagree with you on the signficance of this finding — or at least its practical usefulness.

It is hard to get people to engage on your terms, even if your terms are the most accurate ones. They are inclined to engage on their own. If their own findings contradict their own beliefs, they are more likely to assess their beliefes than if your more robust findings do — even if their methodology is flawed.

Convincing someone to revisit their own thinking is much harder if you only use their own thinking to help them to do so. It is simply too easy for people to dismiss others’ thinking.

****************

There is a good reason why pen and paper tests are less accurate at the extremes, relative to their accuracy — or at least reliability — in the middle. Simply put, more questions at that difficulty level make it more reliable at that level. Tests are geared towards where most of the students are expected to cluster. Think of the bell curve, right?

Making it accurate further out on either end requires a lot more questions, questions that are not going to contribute information about most students’ abilities.

The promise of computer adaptive testing is that it can give students questions clustered around own abillitiy levels, without requiring them to answer lots of other questions. For now, however, we should already expect that NYS tests will be more accurate around the 2/3 line than around the 1/2 or 3/4 lines.

Of course, you already know that, right?

I’m sure there’s a way around it, and maybe it starts with limiting such an analysis to middle and high schools.

Caroline Hoxby’s recent study indicates that the vast majority of charter school students in New York City enter charter schools in kindergarten or first grade. This means that they are not transferring from parochial schools. But it also means that there frequently is no paper trail on the students entering charter schools that would allow us to distinguish between the selection and influence scenarios.

You state “Longitudinal data on individual students would be the best way to examine this distinction, but in the absence of that, we might look at the percentages of students identified as ELL or special education at the time they enter a charter school and compare that to the percentages of ELL and special ed students at the same grade level in traditional public schools.”

One important point – charters don’t only draw from traditional public schools but also parochial schools in which students don’t have IEPs, etc. – which means intake percentages might be different. It also doesn’t address the reason why IEP/LEP numbers decrease in charter schools – opponents say that it’s because the schools force those kids out while the schools tout their ability to remove kids from the lists of the labeled.

Peer schools are defined in terms of students’ proficiency on 8th-grade ELA and math scores, the percentage of special education students, and the percentage of students overage for grade. School size and the school’s method of selecting students don’t figure into the calculation. It’s assumed, I think, that the matching of schools on the basis of 8th-grade test scores will adjust for differences in the screening that high schools use in admitting students. I *have* heard high school principals express concern that their peer schools may not be matched fully on resources, in that some peer high schools may (in spite of Fair Student Funding) have much higher per-pupil expenditures than the school itself.

I agree with KS that charter schools are likely doing things that are precluding needing IEPs. We have students enter our charter school from the district that have completely ridiculous IEPs. We try to get them off of their IEPs if we find that they are unnecessary.

Secondly, because schools get extra $$ for enrolling ELL students, there is an incentive for schools to label kids and keep them on the ELL rolls. We find that most ELL students that come from the district to our school no longer require labeling, or can be unlabeled relatively quickly.

That being said, charter schools are often too small to adequately serve the most severe special ed cases. New immigrant families are also definitely harder to attract to a young institution that people who are already frustrated with the larger system.

The data points, I think we all agree, are ethnicity, poverty, disability status and ELL status.

The first two have very little directly to do with the school, but the last two have a whole lot to do with the school. Let me speculate.

On ELLs:

(1) I believe there are relatively few charter high schools. I don’t know why that is. I’m going to guess that high school ELL percentages, however, are higher citywide due to the arrival of recent immigrants than the percentages in elementary and middle schools. Is it possible that a fair comparison between the grades of charter schools and the grades of the TPS (to borrow from CREDO) would yield a smaller ELL gap?

(2) Since many charters start in kindergarten, might they have particularly effective sheltered English or ESL programs and thereby reduce the number of ELL students?

These are only hypotheses for why charter ELL figures might be lower than that of the general district population. You pointed out a third (neighborhood of location, easily testable enough by comparing the closest zone school’s ELL rate to that of each charter). Again, I’m not saying one way or the other, but I’d like to see these issues studied.

However, at first glance, the figures of 16% and 4% seem to indicate that there is a huge disparity and charters are under-serving ELL students. I’d like to see my caveats explored by the wonks out there.

On students with disabilities:

I think we are a long way from seeing the full sunshine and light of day on the practices of district schools when it comes to IDEA, RTI and referrals to the CSE. I personally believe, from my own experience, that over-referral to the CSE in poor communities is rampant, and that some or maybe even many (I can’t say all) charters support students in ways that preclude needing an IEP, unnaturally driving down the special ed enrollent percentages, but not changing the makeup or “at risk” nature of the population.

BTW the charter school law requires charter schools to make “good faith efforts” to enroll comparable numbers of Special Education and ELL students as district schools. There is legislation pending to require them to do so … the DOE report may give a boost to the legislation.

A correction: small schools and large schools fall under the same student selection/placement rules. (some exceptions, small numbers of schools require exams or special talents, i.e., Stuyvesant, Performing Arts, etc.)