The Department of Education didn’t want you to read this story
As reported in the NYTimes over the weekend, the Department of Education released a report on Friday from a study conducted by the Educational Testing Service, comparing test scores of public and private school students. While private school students have better scores, on average, the study found that after you control for various school and student characteristics there was no difference between the test scores of public and private school students.
This doesn’t really tell a parent trying to pick a school very much. Among the things controlled for were school size and the composition of the teaching staff, which are precisely the sorts of things that fancy private schools pride themselves on. And the schools that don’t offer these advantages aren’t necessarily competing on better educational offerings, but on safety (e.g. that they can kick the troublemakers out) and values. But it does complicate matters for the privatization advocates who argue that private schools are going to do a much better job with the same population of students and the same resources as public schools.
The most entertaining part of the story is the Education Department’s attempt to claim that releasing the report on a summer Friday, without a press conference, wasn’t an attempt to bury it. Yes, this may have been the first day it could have been released. But if the results had been favorable to private schools, I’m confident that it would have been released in an event at a school where at least some students are attending through vouchers.
July 18th, 2006 at 10:35 am
Hi Elizabeth:
I think the important result from this analysis (I’ve read the first half now) is that the difference in scores on standardized tests can be largely explained by _student_ characteristics, which include gender, race/ethnicity, students with disabilities, english language learners, and a list of other characteristics listed in Figure 1-1.
Including school size & teacher experience (which you cite here) were part of the “school-level” variables, further diminishes the difference between public/private schools. (table 2.1-2.2), I think, but it’s not the main driver of the effect. (There’s also overlap between the school-level/student level variables, which I haven’t read about deeply enough to understand).
I find the study compelling, but then I’ve never believed that privatization adds value in of itself.
They’re mealy mouthed in their sections describing problems of interpretation (which are absolutely correct, but would certainly be just as important if the numbers had worked out the other way).
I hope there’s more discussion of this report even though they tried to bury it. I honestly believe that we are asking for the moon and the stars from the public schools, and they fail at delivering that, but do a solid job with their population of students. When I hear from public school teachers, I’m shocked at the level of services teachers and schools are routinely providing, with no fanfare. The most recent was a story of a school secretary who was taking on the responsibility of personal care (brushing hair, washing face) and providing breakfast to a student whose family wasn’t doing it. Another is from the blog “the life that chose me” which describes a teacher taking on the role of nurse and personal caretaker.
bj
July 18th, 2006 at 10:38 am
oops, I jumped the gun. I need to read the report more closely. The school variables were included as level two covariates (and I was ignoring that when I said the school variables weren’t included).
Ignore me while I reasess.
bj
PS: but, I’d certainly appreciate more interpreation with others more familiar with HLM (heirarchical linear models).
July 18th, 2006 at 11:34 am
I really hate the emphasis on test scores as a barometer of success. I feel like test scores are only part of the story of any given school. I guess they get a lot of play because they’re so easy to measure and compare, but I don’t think they’re very useful.
At the local level, test scores only tell a parent trying to pick a school so much. Besides, a parent trying to pick a school isn’t (or, at least, I would hope they wouldn’t be) interested in broad sweeping generalizations. There are amazing public schools and amazing private schools (and terrible versions of both, too). Parents should be looking purely at the local level, not at national studies.
The Department of Education, on the other hand, has to be thinking nationally, and apparently is looking for data to guide their decisions. (Or at least to apppear to justify the decisions they’ve already made!) I guess you have to start somewhere with data, but I still think this report should have very little to do with education policy – in any direction. Kind of a yawner. Maybe that’s why it slipped out with no fanfare…
July 18th, 2006 at 1:26 pm
I was pretty much off base (the two sets of variables they model are ones that can be assigned to an individual (i.e. is the student male/female) and to a school (i.e. the percent of certified teachers). They use both in their assessments, and the bottom line conclusion is that if you control for all the variables, the difference in test scores between private/public schools disappears.
I would like to see assessment of the impacts of each of the individual factors (but, presumably the sample sizes of private schools weren’t big enough to do it).
But, I think that thinking about the report as a way of telling you what to do with your own kids isn’t the right role for the report.
July 18th, 2006 at 2:20 pm
I don’t think the purpose of the report is to help individual parents. I think they expected it to show a very different result, which would then be justification/rationalization for diverting public moneys into private schools, especially “Christian” schools. But that backfired, since the report I read is that the “Christian” schools came in dead last, substantially worse than public schools. And if public schools do just as well overall as private schools, all else being equal, then there’s no argument that private schools will serve students better. So the money can go to education rather than into the pockets of private investors or mega-churches.
I think the bottom line here is that public schools are doing as good of a job as private ones as a whole, especially given the fact that they have to serve the entire population and cannot cherry pick only those that they want to have in their schools. It’s important because it takes away one of the major arguments of the right-wingers who want to privatize public education.
July 18th, 2006 at 2:49 pm
I am finding it hilarious (on my good days) and incredibly depressing (on every other day) watching the data repeatedly prove the Bush administration wrong. WMD? Er, never mind. Global warming? Oh yeah, I guess that actually is happening. Christian schooling? Not only not better, but actually worse. Although you’ll notice the Lutherans fared quite well!! Once again showing that people from Minnesota are all above average. (But then again I’m biased.) 😉
July 18th, 2006 at 9:31 pm
I was just at a meeting where some high school kids from a large urban high school in an economically depressed area were complaining about the effect that ‘test prep’ classes have on the rest of their academic career. Basically, the school was requiring that kids go to test prep at the expense of their real classes, so that the school’s test results would look good. Of course, the kids, who were missing academic classes, were in danger of failing.
But since the DOE is all focused on testing now, it’s ignoring the negative impact on individual students.
July 19th, 2006 at 8:42 am
Just as we all suspected… They hoped this report would support the policy they are pushing, announced yesterday, and reported on in today’s NYTimes: National privatization of public education:
http://www.nytimes.com/2006/07/19/education/19voucher.html
Republicans Propose National School Voucher Program
By DIANA JEAN SCHEMO
Published: July 19, 2006
WASHINGTON, July 18 — With Education Secretary Margaret Spellings joining them in a show of support, Congressional Republicans proposed Tuesday to spend $100 million on vouchers for low-income students in chronically failing public schools around the country to attend private and religious schools. (more at link)