Posted by Barnett Berry on Thursday, 05/15/2014

The Center for Teaching Quality

It’s a bad narrative. But it stays at the surface. It’s easy to message. It doesn’t require nuance or complexity of thought. (Thank goodness! Whew!)

And so it abides, the school reformer’s logic… Teachers play an important role in student achievement. So if a school is low-performing, it’s because they’ve made the mistake of hiring “bad” teachers (probably protected by unions). Good teaching is about being smart and caring, so just find people like that and give them a month or two of training. Oh, and experience doesn’t matter much, so there’s no point in worrying too much about it. 

Sadly, many fall for this myth of teaching made simple—and cite research to defend it.

THE RESEARCH ON EXPERIENCE

Several recent reports claim that after a teacher’s first two years, additional teaching experience doesn’t matter for student achievement. But these studies, produced by researchers like Eric Hanushek, generally refrain from acknowledging counter-evidence, like the extensive studies conducted and reviewed by Jennifer Rice.

And they neglect the warnings of Harvard economist Dick Murnane, who pointed out decades ago that the answer to the “effects of teaching experience on student achievement" question depends on whether a study uses longitudinal or cross-sectional databases.

Translation: it’s one thing to analyze student test scores over time of seventh-year teachers who’ve taught the same subject and grade each year. It’s a different matter to gather one year’s student test scores of seventh-year teachers regardless of whether they have ever taught the grade or subject before. Hanushek did the latter

Contrast Hanushek’s results with those of Ben Ost, who took the longitudinal approach. He found that teaching experience actually matters a great deal for student achievement far beyond the first two years, if teachers are assigned the same grade level and subject each year.

And to consider the correlation between teacher experience and achievement at the school level—Tony Bryk and Barbara Schneider have shown that schools improve over time when there’s relational trust among parents, teachers, and administrators. That kind of trust? It takes time to cultivate. A school with transient teachers—only there for a year or two—isn’t likely to build it. 

NEW FINDINGS THAT DESERVE MORE ATTENTION

And now Sunny Ladd of Duke University has released a study that confronts the recent reform wisdom head-on. Using a database that captured test scores of 1.2 million students, Ladd found that teaching experience matters for at least 12 years of teachers’ careers. This held true for middle school teachers of math and those of English and language arts.

While the returns level off after 12 years, Ladd notes that math teachers with 21-27 years of experience “were still about 85 percent more effective than they would have been after two years of experience and about 30 percent more effective than after five years.”

Perhaps most noteworthy is Ladd’s finding that teaching experience matters a great deal when it comes to reducing student absenteeism.  Based on her calculations, novices have a very limited effect on this important non-cognitive outcome regardless of quality, but “a teacher of given quality, who obtains over 21 years of experience on average, reduces student absences by 8 days.”

THE SIMPLICITY THE PROFESSION NEEDS

Don’t get me wrong: Teaching experience, in and of itself, does not matter for student achievement. The effects of teaching experience hinge on a teacher’s professional development, the types of classrooms he or she is assigned, and the conditions under which he or she works.

But today’s reformers need to get brave, absorb the research about why teaching is complicated work that takes years to master, and rethink policies that devalue classroom experience.

Maybe the truth about teaching actually is simple: It’s complex and important work, so making (or writing about) the profession’s policies requires us look to research and do so critically.