Education report gambles big bucks on skinny data

Posted on February 16, 2005 | Type: In the News | Author: Robert Robb
  • Twitter
  • Facebook
  • Email

The Rodel Foundation says that if Arizona spent $1.7 billion a year on particular programs, "it is reasonable to expect student performance to double within 10 years."

That's an audacious claim. It's also, as it turns out, a claim utterly without foundation.

Rodel's report, "Lead with Five," and its backup technical paper provide considerable detail about its cost estimates and their methodology. But not a word of explanation or methodology about the finding that these particular investments will double student achievement in just 10 years. There's not even a clear statement about what particular measure of student achievement should reasonably be expected to double.

For elucidation, I turned to Carol Peck, the foundation's president. She didn't have a ready answer, and referred me to Allan Odden, the report's consultant.

Odden was very nice, informative and patient. But only slightly more illuminating about the claim and its justification. Probably, he said, the reference was to doubling the percentage of Arizona students scoring proficient on the federally administered National Assessment of Educational Progress tests.

Now, that would be remarkable.

In its report, Rodel cites Arizona's lagging NAEP performance in reading and math in the fourth and eighth grades. But right now, there are no states with twice Arizona's percentage of students ranked as proficient in fourth-grade reading or math, or eighth-grade reading. And only one, Minnesota, in eighth-grade math.

So, that would be a very long way to go in a relatively short time, at least in public policy years.

Rodel recommends that the money be invested in five areas: all-day kindergarten, teacher development and merit pay, smaller schools, smaller class sizes in the early grades, and tutoring.

Rodel says its recommendations are "research based." But it offers no specific estimate of student achievement gains that could be expected from them singly, or how cumulatively they would add up to a doubling of student performance in 10 years.

Moreover, it skips blithely over some knotty problems in the research and its application to Arizona.

Take the current controversy over all-day kindergarten. While most researchers accept that all-day kindergarten provides short-term gains in student learning, there is considerable doubt about whether those gains are or can be sustained in later grades.

The case for skepticism is well made in a recent Goldwater Institute report by its president Darcy Olsen, formerly an early-education specialist for the Cato Institute.

But assume the gains are real, particularly for low-income and minority children, as Rodel claims. In Arizona such children are, for the most part, already in all-day kindergarten, funded at the local school district level. Rodel never discusses why gains in student learning should result from transferring this cost to the state.

Rodel bases its claims on smaller K-3 classes primarily on an experiment done in Tennessee during the 1980s, in which students were enrolled in different-size classes and kept there from kindergarten through third grade. Students in the smaller classes on average did better on achievement tests.

As economist Eric Hanushek has pointed out, however, there are some problems generalizing from the Tennessee experiment. In about half the schools, there was no improvement reported from the smaller classes. Moreover, all the gains were in the kindergarten year.

Broader studies consistently show little correlation, if any, between teacher-student ratios and student achievement.

And for what it's worth, Tennessee ranks in Arizona territory on the NAEP tests.

I think merit pay is a good idea, but I haven't seen any solid research that it improves student performance. Odden said that's because it hasn't really been consistently tried. A fair point, but then how can increased student learning be projected based on it?

Smaller schools have benefits, but primarily in improved student behavior and reduced dropout rates in higher grades. Arizona has already built behemoth schools, so Rodel recommends creating schools within schools.

A worthy experiment, but Odden says the research on that, as opposed to truly smaller schools, is sketchy.

Odden in conversation shows more academic circumspection than the Rodel report exhibits in print. Circumspection is warranted in these education debates. Right now, the favored reforms of both the left (more money for educational inputs) and the right (choice and competition through charter schools and vouchers) have demonstrated, at best, improvement on the margins.

Rodel's claim that student achievement in Arizona can be doubled in 10 years if enough money is spent on the right programs isn't really research-based.

It's the education establishment's version of a faith-based initiative.

Reach Robb at or (602) 444-8472. His column appears Sundays, Wednesdays and Fridays.

Advanced Search

to Go >>

Recent Facebook Activity