Monday, February 20, 2006

Voucher advocate proposes "no fiddling with the data."

Another recent study of 13,577 schools shows that public schools hold their own against private schools. The report calls the public school students’ scores “statistically significantly higher,” because 10 points is generally considered to represent one grade level’s difference. “This analysis makes it appear that there isn’t anything magical in private schools that leads to a significant difference in achievement,” Mr. Lubienski, co-author of the study, said.

In what has to be one of the most nonsensical quotes ever, Joe McTighe, Executive Director of the Council for American Private Education had this response to Education Week, “When you look at the data as is without any fiddling, students in privates outperform those in publics by a wide margin,” he said.

Let me interpret "without any fiddling." It means "don't look too closely at the data." When they broke the data down and compared students with similar backgrounds, public schools outperformed private schools.

Look for a huge push from voucher advocates in the next year. Indiana has already vowed to do so. I think there will be more desperation to do so soon because the longer time goes on the better and more thorough the research base will get. And that....might not be what private schools and voucher advocates really want.

2 Comments:

At Monday, February 20, 2006, Anonymous Anonymous said...

No, Private Schools Aren’t Worse

By Greg Forster

Feb 10, 2006

Opponents of school choice have just released two major studies claiming to show that public schools actually perform better than private schools. One study made the front page of the New York Times. Unfortunately, both of them are seriously flawed. What’s more, a much larger body of much better studies reaches the opposite conclusion.

The studies were released within about a week of each other, and by the same organization: the National Center for the Study of Privatization in Education. While this may have made it a good week for the center in terms of publicity, it was an awful week in terms of scientific credibility.

The first study found that when you control for demographic factors like race and income, public school students actually have higher test scores than private school students. The authors have been aggressively touting this study as though it showed that public schools were better than private schools.

In fact, it shows nothing of the kind. They take snapshots of student achievement in isolated years rather than tracking achievement over time. That may not sound important, but it’s crucial. If you don’t track students over time, you can’t find out why one student has a higher score than another.

Single-year test scores mostly reflect student quality, not school quality. A student with high test scores is usually just a good student. It takes a student whose test scores are rising to prove that the school is good.

A much more likely explanation for these data is that the students who enter private schools tend to be slightly worse students than those of the same race and income who enter public schools. That makes perfect sense, because within each racial and socioeconomic group it’s the low performers whose parents will want to make the sacrifices necessary to put them in private schools.

The other study looked at Cleveland’s ten-year-old voucher program. Using a new statistical model to analyze a previously existing data set, it finds that kids remaining in Cleveland public schools do better than voucher users.

This study is even more flawed than the previous one. The data set it analyzes does not allow for valid comparisons between similar student populations. The voucher students and the public school comparison groups in the data set are dissimilar not only because one group uses vouchers and the other doesn’t, but also in a host of other ways, and there’s no way to disentangle what’s really causing the test score difference. The study compares apples and oranges.

As it happens, numerous studies that avoid these methodological problems find that private schools do better. Most convincing are seven studies that compared students who won a random lottery to use a school voucher at a private school to similar students who lost the lottery and stayed in public schools. All seven found that voucher kids did better. Studies using other methods also favor private schools overwhelmingly.

Not all of the work sponsored by this organization is as bad as these two studies. It has promoted good, solid research showing that competition from school choice improves public schools. That’s even more impressive given that it’s housed at Columbia University’s Teachers College, the greatest academic stronghold of the teacher unions anywhere in the nation.

But the center also has a dark side. From time to time it will release badly flawed studies purporting to show the inferior performance of private schools. With these two studies, it has just had probably its worst week ever.

Let’s hope this misleading research doesn’t distract from the real scholarly consensus finding that private schools do better, and that school choice works.


Greg Forster is a senior fellow at the Milton and Rose D. Friedman Foundation.

 
At Monday, February 20, 2006, Anonymous Anonymous said...

And for more debunking of this ridiculous "study", read here:

http://www.edexcellence.net/foundation/gadfly/issue.cfm?id=227#2668

 

Post a Comment

<< Home