Brigham Young University Homepage



News Release

Study answers how much last year’s teacher affects this year’s test scores

The impact a specific schoolteacher has on students’ math and reading scores – whether bad or good – fades quite fast, according to a new study by researchers at Brigham Young University and the University of Michigan.

A recent trend in public education is to measure teacher quality based on how the students fared on standardized tests compared to previous years. If most of Mr. Green’s current 5th grade students score at a higher percentile than they did as 4th graders, then Mr. Green gets what’s called a high “value-added” rating.

In August, The Los Angeles Times shined the spotlight on this approach by publishing rankings for 6,000 L.A. schoolteachers based on value-added analysis.

The new study instead measured whether teachers like Mr. Green put students on a higher trajectory in the years to come. The researchers report that most of the gains from a highly rated teacher vanish quickly. In reading, 87 percent of the benefit fades after one year. In math, 73 percent of the gains fade after one year.

“People are looking for a silver bullet to fix public education,” said BYU economics professor Lars Lefgren. “We’ve shown that the benefits are mostly transitory, so you don’t want to sacrifice everything else you might value in a teacher just for value added to test scores.”

Lefgren and fellow BYU economist David Sims co-authored the study with the University of Michigan’s Brian Jacob for The Journal of Human Resources. Their analysis included eight years of data from 1.3 million student test scores in North Carolina schools.

While the news may sound depressing, the report offers this silver lining: The effect of having a crummy teacher doesn’t last long either.

“You probably won’t be scarred from having an incompetent teacher like me,” Lefgren joked.


Follow BYU news on Twitter: @BYU


Report Abuse

Login to post a comment.BYU students and employees can comment on this story by logging in. Comments are moderated and will remain posted if they are on-topic, use clean language and show respect for others.

Like Union, Like Student, eh?
  12/7/2010 11:53 PM by James

JayR: I grew up in both public and private schools. I was blessed in both settings When i learned more and more about the unions, i saw how great teachers are sometimes kept out so that poor quality teachers can stay in; some tests are written in a way to hide the quality of a given instructor; teacher's union members argue how their system is just fine the way it is, and if it wasn't, then a new system would have come along years ago (unless the school unions were against a new educational system?).

Bob Bowden from <a href="">The Cartel Movie</a> says:

"So hubris and failure, it seems, do occasionally make a joint appearance. And we can usually laugh it off. But there's a troubling difference when adults practice the education version of this combination: They pass it on to students. As a life lesson. Because it serves their interests.

"This pernicious practice was first famously reported in Charles Sykes' 1996 book, Dumbing Down Our Kids: Why American Children Feel Good About Themselves But Can't Read, Write, or Add. More recently, Davis Guggenheim's Waiting for 'Superman' reiterates the same idea. Today, U.S. kids rank 25th in the world in math, and first in confidence about test performance.

Like union, like student, eh?

"So when I hear excuses from defenders of the establishment, I can't help thinking to myself, don't they know what the rest of us are thinking? Don't they know how we're finishing their sentences for them?

"There will be the public part of their statement, then the other part they believe, but intentionally leave out.

* We wanted to be treated like professionals, but we want job guarantees that no other professionals have.

* We support teacher accountability, provided it virtually never results in actual dismissals for tenured teachers.

* We should implement policies to prevent burnout of young, hard-working teachers, as long as the most talented, hardest-working teachers are never individually rewarded for the exceptional efforts.

* We believe in developing fair methodologies of teacher evaluation, but in practice we'll declare that any particular evaluation plan is unfair.

* We all want to reform public schools, provided reform is defined as "good for jobs."

* Not enough money reaches the classroom, but we won't publicly criticize excess administration spending because one of those jobs might be ours some day.

* We all want better schools, provided that "better for kids" and "better for adults" are never distinguished." ~Bob Bowden, director of <a href="">the Cartel Movie</a>

Student Feedback
  12/1/2010 12:41 PM by Eric

What would happen to public education if teachers were evaluated carefully according to (blind) student evaluations and surprise peer evaluations, the way that BYU professors are evaluated? I have seen through my children that many teachers in public schools do not often adapt to student feedback. More careful comprehensive evaluation could close the feedback loop and provide opportunity for improvement. Proper incentives for improvement would grease the wheels.

A Few Thoughts
  11/30/2010 10:02 AM by Kimberlee

There is an enormous amount to think about here, but below are a couple thoughts:

1. This only refers to the persistence of a teacher's impact on test scores. As mentioned, a teacher may have longer term impact in other ways.

2. The results, as far as I know, suggest average effects for all the teacher's students do not persist. A great teacher may have significant impact on one or a few students that persists indefinitely. Likewise, a poor teacher may do harm to a child that persists for a very long time. I would be particularly interested in whether there is more or less persistence within specific demographic groups.

3. I wonder how the persistence relates to a student's future teachers' impacts. It would seem that, if a teacher the following year has great impact on the student, that would mitigate the impact of a teacher the previous year. If the teacher in the following year, on the other hand, had a more neutral effect on the student, then I would expect the previous teacher's effect to persist a little more. What I mean is that if the year x+1 teacher effect is large, that would mean the year x teacher's effect on the student's trend would HAVE to have not fully persisted. If I'm missing something here, I'd love to hear it.

But, thanks Dr. Lefgren for what you have added to this debate.

a very bad teacher can affect an entire life
  11/30/2010 5:37 AM by Fauna

I had a son who was reading very well before the second grade. He was sounding out words like Nebuchadnezzer and progressing quite well. Then his second grade teacher started criticizing his mishaps in reading and telling him that he would flunk if he could not improve. The children in his class would repeat her negativism towards him on the playground. One day I was called in to school when 5 of his classmates decided to beat him up together. He survived physically, but still has major reading problems at 23. However, his reading in the scriptures is at a much higher level than his other scholastic attempts. I would surmise that he opens up more easily to his successes than to her attempts to assure his failure. If there is a way to weed out teachers such as her I am all for it, as I can see lifelong damage in this case. Exception to the rule, maybe. Affecting this generation and many generations to come, absolutely.

  11/25/2010 11:54 AM by Scott

The purpose of the value-added approach is to measure the quality of the teachers; this study does not undermine the ability of such approach to determine the quality of teachers, it only undermines the value of quality teachers. I bet the effect of having NO teacher would fade after a few years as well.

Go Lars!
  11/22/2010 10:57 AM by Benjamin

Prof Lefgren is a stud. Enough said.

What does qualitative data show?
  11/19/2010 11:17 AM by Carrie

Sounds like the study could use qualitative triangulation. I know of lots of cases--some personal--where a "good" teacher's effect lasted years, while a "bad" teacher turned students off the subject.

Also, we have to keep looking for ways to make teachers accountable. While many are very competent, many aren't--and most of what a teacher does is unobserved, except by their students.

RE: Consecutive good/bad teachers
  11/19/2010 9:44 AM by Joseph

That's something I asked Prof. Lefgren in the interview. Because the benefit on test scores fades so quickly, the effect of consecutive high value-added teachers is not nearly as strong as many people assumed.

Some prior literature stated that 5 straight high value-added teachers could possibly be enough to close the achievement gap for students from low-income households. This study, which followed students from 3rd grade through 6th grade, indicates that some of the optimism surrounding value-added ratings should be tempered.

What about the effect of having good/bad teachers over mutliple years?
  11/19/2010 8:05 AM by Bryce

True, a single year in a classroom of a "highly rated" teacher isn't likely to have much of a lasting impact. But, what about students who have a high-performing teacher for multiple years (or, sadly, are in a school environment where they suffer through multiple years of poor instruction)?

Studies published in the Education Trust seem to suggest that multiple years in the classrooms of teachers who have high value-added ratings does make a difference. H

How about a multi-year study designed like the one above? I'm a little worried that the present findings can be used by underperforming schools & districts to justify their lackluster attempts to improve teacher performance.

Story Highlights

  • The analysis included eight years of data from 1.3 million student test scores in North Carolina schools.
  • Most of the gains from a highly rated teacher vanish quickly.
  • The effect of having a crummy teacher doesn’t last long either.

Click here to download
Study finds that the effects of teaching, good and bad, fade quickly.
The Church of Jesus Christ of Latter-day Saints | BYU-Hawaii | BYU-Idaho | BYU Jerusalem Center | BYU Salt Lake Center | LDS Business College | Missionary Training Center
Updated daily by the BYU Web Team Brigham Young University Provo, UT 84602 (801) 422-4636 Copyright © 2009. Brigham Young University. All Rights Reserved.