If you listen to a lot of policy discussions on education, chances are that you've heard one scholar or another stand up to talk about how teacher credentials, such as holding a traditional license or having earned a master's degree, don't seem to matter much when it comes to improving student achievement.
Duke University researcher Helen F. Ladd says that there are two problems with those studies. The studies are: 1) old, and 2) focused mostly on elementary school children.
To gather newer data on the impact of teacher credentials and characteristics on high school students' achievement, Ladd and her research partners took a look at scores from the end-of-course exams that all high school students are required to take in North Carolina. They looked in particular at statewide data for four cohorts of 9th and 10th graders for whom they could find and match up data on their teachers. (The final sample included tens of thousands of students.)The bottom line, the researchers found, was that at the high school level, most measurable teacher credentials do indeed matter. And they have a large enough impact on student achievement, Ladd and her colleagues say, to suggest that they ought to figure into policymakers' decisions on how to raise the quality of instruction in schools.
In keeping with previous studies on teacher quality, the North Carolina data show that teaching experience matters—up to a point. After five years on the job, another year of experience didn't seem to make that much more difference. The researchers also found that teachers who had graduated from more-selective colleges spurred bigger learning gains in students than those from less-selective schools.
With regard to master's degrees, the researchers' findings were a bit more nuanced. Teachers who had earned a master's degree before entering the field were no more effective than those without master's degrees. But teachers who got a master's degree after they began teaching were found to do a better job at boosting students' test scores than did their less-educated teaching peers.
Getting a high score on the subject-matter tests that teachers take for certification also was linked to greater student learning gains—especially in algebra and geometry. Likewise, teachers who were certified in the subject they taught were found to be more effective than those who were not.
The study also found that teachers with a "lateral"—or alternative—license were slightly less effective than teachers with traditional teaching licenses.
Teachers' earning certification from the National Board for Professional Teaching Standards also helped boost students' scores. But what's worth noting here is that the learning gains begin to show up in the year during which teachers are making their applications, which suggests that the process itself may improve teaching.
Now for the bad news: The researchers found, as have previous studies in other states, that teachers who had the "right" test scores and credentials were unevenly distributed among schools. Schools with high concentrations of poverty were more likely than schools serving wealthier populations to have alternatively certified teachers and novices. They were less likely to have teachers with high test scores or degrees from competitive schools. And black males taking Algebra I were about 22 percent more likely than white females to be taught by a beginner.
All of this is worth keeping in mind because many experts are calling for judging teaching effectiveness based mostly on students' test scores. While based on data for just one state, these results suggest other factors may useful markers of teacher quality, too. And It's possible, these researchers say, that the end-of-course exam scores used for this study may actually be a better barometer of what goes on in a classroom than the broader exams that students take in earlier grades.
Finally, the researchers turned up some unexpected findings that puzzled and disturbed them. The data showed that black teachers teaching white students and male teachers teaching female students were linked to negative effects on student achievement.
"We find those results quite distressing and worthy both of more research and, assuming they hold up in other studies, more attention by school officials," Ladd writes in an e-mail response to my query on that last point. "We would not recommend rearranging teachers and students to avoid those pairings, but rather would support efforts to minimize the adverse effects in the future."
The full study is in the current issue of the Journal of Human Resources, which is a subscriber-only publication. But an earlier version of the study can also be found on the website for the National Center for Analysis of Longitudinal Data in Education Research, better known as CALDER.