The U.S. Department of Education recently released a study comparing the performance of public school students with students at private schools. Reg Weaver, president of the National Education Association, concludes from this study that public schools are “doing an outstanding job.”

A careful examination of the study’s findings shows, however, that the accurate conclusion is that the longer many students stay in public schools, the worse they do. The federal study tapped student test score data from the 2003 National Assessment of Educational Progress test and focused on fourth- and eighth-grade math and reading scores.

The spin propagated by teachers’ unions includes the assertion that public school students are doing as well or better than their private school peers in most categories of education. Their observations conveniently fail to acknowledge the obvious trends in the data.

For example, the study does show that public school fourth-graders performed as well as private schoolers in reading, when race, gender and other factors were taken into consideration. However, by eighth grade, private school students consistently scored higher than public school eighth graders. A similar trend appears in math.

Fourth-grade math scores show that public school students scored higher than private school students, but both groups performed at about the same level in eighth grade. Thus, in reading and math, the trend appears to be downward for public school students vis-à-vis students at private schools. This result mirrors trends seen in international comparisons of student achievement.

The teacher unions should also beware of putting all their eggs in the federal study’s basket because of its limitations. First, the study uses test data from only a single year. This snapshot technique doesn’t allow for analysis of long-term student improvement or decline.

Thus, for example, while fourth-grade public school students may outscore private school fourth-graders in reading, we don’t know the ability level that these students started off at in the first grade. Public schools may have improved the performance of their students, or it may be that, as the federal study cautions, “Apparent differences in average achievement between public school students and private school students may simply reflect average differences in achievement between their respective student populations.”

Further, the study wasn’t conducted as a scientific experiment with some students randomly placed in public schools with others randomly placed in private schools. The study admits that this deficiency opens up the possibility of so-called “selection bias,” which means that differences in student performance may be explained by factors other than whether a child attended either a public or private school.

Finally, private schools differ not only from public schools, which tend to be more homogenous, but from each other. A Montessori school is much different than a Catholic school. Lumping all private schools together, therefore, fails to enlighten the private-public comparison.

The authors of the federal study admit that the study has only “modest utility.” It does, however, reinforce what we already know — too many of our public middle schools have become the gateways to student failure. Such underdevelopment isn’t something Reg Weaver and his fellow union members should brag about.