THESE TEACHERS GIVE KIDS, EDUCATORS FOOD FOR THOUGHT

Saturday, June 26, 2004

How effective traditional teacher-training programs are, or whether they're even necessary, is a controversial question, pushed into the forefront by the requirement of the No Child Left Behind Act that every teacher be "qualified."

There's very little empirical evidence linking student achievement with teacher qualifications, though what there is suggests that certification is no guarantee of competence, let alone excellence. And research is difficult, first because parents tend to object when their children are enrolled in experiments to find out what helps them learn and, second, because there's usually no way to do an experiment blind, as would be standard in clinical trials for medical treatments. If the idea is to compare math curriculums, for instance, participants have to know what curriculum they're using.

But an intriguing natural experiment has been occurring, in the form of the program Teach For America, which was started in 1989. TFA recruits recent college graduates who have demonstrated both leadership and high academic performance, puts them through an intensive five-week training session and places them in low-income schools.

Three foundations active in education commissioned Mathematica Policy Research Inc. of Princeton, N.J., to carry out a comparative study of TFA teachers and other teachers in the same schools and the same grades (www.mathematica-mpr.com). In math, the students taught by someone trained by TFA made bigger gains over the course of the year than those taught by non-TFA teachers, and the difference was statistically significant. In reading, the gains were essentially the same.

The study included 17 schools in six districts, 100 classrooms and nearly 2,000 students. The researchers limited their project to grades one through five, because it was easier to separate TFA teachers (anyone who had entered teaching through the program) with a control group consisting of everybody else -- that is, some of the teachers in the control group had regular certifications, some had alternative certifications and some had no certification at all.

If it seems unfair to include noncertified teachers in the control group, remember that most of the TFA teachers aren't certified when they enter the classroom either, though many do earn certification in their first year or two.

The study compared TFA teachers both to the entire control group and to just that part of it consisting of novice teachers, in their first three years of teaching. The difference in math progress was larger than for the control group as a whole, and there was still essentially no difference in reading.

TFA is very choosy about who it accepts. In 2003, it had 15,706 applicants and 1,656 new corps members. The result is that 70 percent of TFA teachers earned their bachelor's degree from a college with competitive admissions, while for the control group it was only 2.4 percent.

Of course even noncompetitive colleges enroll some excellent students, but it's reasonable to assume that TFA teachers entered college with much higher SAT/ACT scores, on average, than teachers in the control group, especially since it's known that students planning to get a degree in education have lower scores, again on average, than students planning for most other majors.

That matters, because it's been known since the mid-'60s that teachers' own performance on tests of verbal proficiency, such as the SAT, is one of the best predictors of their students' performance, much better than whether they're certified or not. Sociologist James Coleman discovered that while doing research on the effects of school segregation.

But it's also been widely ignored, because, as Coleman explained in a lecture I heard him give, black teachers in segregated schools did much worse on the tests he administered than white teachers did, and nobody wanted to go there. Nobody wants to go there now.

Of course this does not mean black teachers are all less competent than white teachers, only that there is intense competition for highly talented black college graduates, and teaching is not necessarily their most attractive opportunity. The same is true for highly talented women, who are less likely to go into teaching than their mothers or grandmothers, because now they have other choices open to them.

There were other differences between TFA teachers and the control group in the study. They were only half as likely to have a degree in education. They all had done student teaching, though not for a long period, while more than half of the teachers in the novice control had done none. That wouldn't be true if they compared TFA teachers with the entire teaching profession, but TFA deliberately places its teachers in high-need schools that have a large percentage of children from low-income families, and such schools often have difficulty recruiting or retaining experienced teachers.

In short, TFA offers districts a way of increasing their pool of well-qualified applicants who can deliver improved student performance. It doubled in size from 2000 to 2003, and in the fall it will have teachers in 22 districts. A lot of poor children stand to benefit.