Best and Worst Teachers Can Be Flagged Early, Says Study
Educators' rankings don't move much
New teachers become much more effective with a few years of classroom experience, but a working paper by a team of researchers suggests the most—and least—effective elementary teachers show their colors at the very start of their careers.
"This is a fundamentally different time period for teachers, when we know they are going through changes," said lead author Allison Atteberry, a research associate in the Center on Education Policy and Workforce Competitiveness at the University of Virginia in Charlottesville. She discussed preliminary results of the study at a research meeting on K-12 and postsecondary education held by the Washington-based National Center for Analysis of Longitudinal Data in Education Research, known as CALDER, on Feb. 21.
"We know less about how these value-added measures work in the early career," she added.
The study tracked the individual effectiveness of more than 7,600 incoming New York City teachers in mathematics and English/language arts. Each of the teachers taught 4th or 5th grade from 2000 to 2006.
The researchers analyzed teacher records from the New York city and state education departments, along with data on the teachers' students, including achievement-test results in math and English/language arts, gender, ethnicity, home language, poverty, special education status, and absences and suspensions.
While incoming New York City teachers became more effective at improving their students’ mathematics and English/language arts performance in their first few years on the job, new research finds that they’re often still in the same performance quintile after four or five years. Researchers compared the mean effectiveness in the first two years with effectiveness in later years.
Ms. Atteberry's co-authors are Susanna Loeb, the director of Stanford University's Center for Education Policy Analysis, and James H. Wyckoff, an education professor at the University of Virginia. They and Ms. Atteberry are all associated with CALDER.
For the incoming teachers who continued to teach for at least five years, the researchers compared the mean value-added effectiveness at improving student achievement in math and English in their first two years of teaching with their effectiveness for the next three years.
Overall, the teachers improved significantly in their first two years in their value-added score. While more than 36 percent of teachers were rated in the lowest of five levels of effectiveness at the start of their careers, only 12 percent were still rated in that same quintile by their third year of teaching.
However, when teachers at each initial level of effectiveness were tracked individually over time, their growth was much less significant. Compared with other teachers who started at the same time they did, teachers in the lowest 20 percent were still likely to be in the lowest 20 percent three to five years later.
"When you look at teachers who in the future are low-performing, very few of those come from the initially highest quintile of performance, and the same is true in the opposite direction," Ms. Atteberry said. "We see that even more at the high end: Teachers who are initially highest-performing are by far the most likely to be in the highest quintile in the future."
Tim R. Sass, an economics and public-policy research professor at Georgia State University in Atlanta, who was not associated with the study, noted that few teachers perform flawlessly in their first year on the job and that policymakers should be cautious about making quick judgments about new teachers. "Even if you're not going to make lots of mistakes, you will make some," he said.
Mr. Sass also pointed out that most teachers who start out as less effective become at least average after a few years. The mean value-added measure in the first two years of teaching accounts for only about 28 percent of the total difference in teachers' value-added effectiveness scores after five years, Mr. Sass said, and the study did not look at specific pedagogical styles or other aspects of teaching that could explain more of the difference.
Still, he said: "Do first impressions matter? The short answer is yes. Early-career value-added does a pretty good job of predicting later performance."
Based on the New York findings, said Mr. Sass, if a principal did not keep the teachers identified in the bottom 10 percent of effectiveness because of their value-added effectiveness in their first two years, the school would get rid of 30 percent of the teachers who rate as least effective five years out, and would keep all the teachers who would eventually be rated in the top 10 percent.
He did not recommend outright that districts try that, but said it was something to consider.
Mr. Wyckoff, one of the co-authors, said he did not think the study should be used to justify firing teachers.
Steven Glazerman, a senior fellow at Mathematica Policy Research in Princeton, N.J., who has studied value-added measures of teacher effectiveness, said the study both confirms and raises some "interesting questions" about the common understanding that new teachers improve rapidly in their first years.
He cautioned, however, that because the study only uses data on new teachers who continued to teach 4th and 5th grades for at least five years, it left out 95 percent of those new teachers.
"What's hard to know about, but what fascinates me in what's exposed by this paper, is the fact that such a large percentage of teachers cycle in and out of tested grades," Mr. Glazerman said. "That makes it very difficult to study them using value-added measures."
Mr. Wyckoff agreed with Mr. Glazerman that the small number of teachers who stayed in the same grades for all five years was "certainly something we were surprised about."
He and the other authors are conducting a follow-up study to gauge whether that finding was due to normal attrition among early-career teachers or something specific to the tested grades.
Vol. 32, Issue 23, Page 6
Get 10 free stories, e-newsletters, and more!
- Director of College Counseling
- Augusta Preparatory Day School, Martinez, GA
- Director of Information Technology
- Charles E. Smith Jewish Day School, Rockville, MD
- Chief Academic Officer
- Cristo Rey Network, Chicago, IL
- Director of Technology
- St. Paul's School for Girls, Brooklandville, MD
- Director of Auxiliary Programming
- Lovett School, Atlanta, GA