Network News

X My Profile
View More Activity

Posted at 8:36 PM ET, 02/17/2010

Why not link teacher pay to test scores?

By Valerie Strauss

My guest is Lisa Guisbond, a policy analyst for the National Center for Fair and Open Testing, known as FairTest, a Boston-based organization that aims to improve standardized testing practices and evaluations of students, teachers and schools.

By Lisa Guisbond
Have your kids ever gotten an A for work that you, or they, didn’t think was worthwhile? Something like that happened recently with Education Secretary Arne Duncan.

Education historian and New York University Professor Diane Ravitch gave him an A for effectiveness at getting buy-in for linking teacher evaluations to student test scores and a D- for pushing bad ideas. I would forgo the A and lower the grade to an F for pushing ideas that are destructive.

Why destructive? At first blush, rewarding teachers for higher student test scores seems reasonable to many people. The second and third blushes are the problem.

It’s difficult to know whether Duncan is paying attention, but much has been written about the folly of payment for ‘results’ schemes in education and elsewhere.

Duke University Professor Helen Ladd summed up the research: "One theory of action seems to be that holding teachers more accountable for the gain in their students’ test scores will induce them to become better teachers.” Ladd says she’s not aware of “any credible evidence” to support this theory.

But there is a wealth of evidence that No Child Left Behind’s focus on test scores has narrowed what children learn by concentrating teachers’ efforts on what they think will be on the test. This is nothing like the wide, rich and engaging curriculum that most parents want for their children and that keeps kids in school. Emphasizing test scores in teacher evaluations is likely to make more kids tune out or walk out.

Educators also say payment for test scores would erode the collegiality that helps them collaborate to improve learning. Teachers say that what kids achieve in fourth grade, for example, is based on the foundation built by first, second, and third grade teachers. Testing children at the beginning of one school year and again at the end of it to see how much they’ve learned sounds reasonable, but it’s not a true indicator of any one teacher’s skillfulness.

Meanwhile, there’s a full team of other teachers whose subjects aren’t the focus of the accountability system: art, music, gym, social studies, and science, even the lunch room and recess monitors. Some children wouldn’t want to be in school if they didn’t have art or music to look forward to. Confidence built up in those areas can also bolster achievement in the two subjects the politicians seems to care most about.

FairTest and the Forum on Educational Accountability have identified effective ways to improve learning without NCLB-style accountability’s damaging effects. To learn more, see the FEA web site.


Follow my blog all day, every day by bookmarking And for admissions advice, college news and links to campus papers, please check out our new Higher Education page at it!

By Valerie Strauss  | February 17, 2010; 8:36 PM ET
Categories:  Education Secretary Duncan, Lisa Guisbond, No Child Left Behind, Standardized Tests, Teachers  | Tags:  teacher evaluations  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: How to keep kids safe en route to school
Next: The trouble with algebra


One reason is that it can lead to this:
"However, according to an erasure analysis conducted by CTB-McGraw Hill, the vendor that provides Georgia’s statewide tests, results at several hundred Georgia schools showed evidence of erasures well beyond what you would ordinarily expect. The company estimated the odds of such excessive erasures occurring naturally at one in a thousand.

The analysis also found that at 74 schools statewide, including 43 in Atlanta alone, more than a quarter of the classrooms tested showed evidence of erasures well beyond the ordinary."

Posted by: edlharris | February 17, 2010 10:38 AM | Report abuse

More anti-accountability propaganda from the folks at Fair Test. These are the folks who would tell you we could eliminate the achievement gap over night simply by eliminating ALL tests. They're also not opposed to promoting and graduating everyone, regardless of their actual academic performance.

The Progressive Education agenda died of natural cause last century because parents wanted to know how their kids were REALLY doing in school. They become disenfranchised with all the pseudo-academic gobbily-gook when schools tried to pass off portfolios and dioramas as authentic assessments. There's a reason John Dewey's philosophy never went mainstream and Summerhill never really caught on. Parents could smell a skunk and didn't like it.

As far as educational historian Diane Ravitch is concerned, she has something the cool-aid drinkers at Fair Test will never have - CREDIBILITY.

Posted by: phoss1 | February 17, 2010 11:03 AM | Report abuse

A hearty thanks to Ms. Guisbond of FairTest for your astute analysis of Duncan’s plan and the fundamental shortcomings of connecting high stakes testing to teacher pay. To it I would add that these schemes tend to most negatively affect the very students we purport to want to help. A 2009 study by the Government Accountability Office on the effects of No Child Left Behind’s high stakes testing mandates showed that schools with high poverty and high concentrations of students of color are most likely to adopt strategies such as narrowing curriculum and teaching to the test. While the whole system is warped by standardized testing policies, these students receive the brunt of them. We appreciate the Obama Administration’s promise to revamp No Child Left Behind, however, throwing away elements such as the Adequate Yearly Progress provision does not remove the single most corruptive issue of over reliance on standardized testing, in fact, connecting it to teacher pay may ultimately exacerbate the problem. With drop-out rates vastly outpacing that of their white counterparts, poor Latino and African American students are floundering in the public education system. This trend threatens to continue as long as we refuse to look for more thoughtful reform strategies, recognize these students’ experiences and meaningfully connect with their communities.

Amina Luqman-Dawson is Senior Policy Strategist at Justice Matters, an Oakland, California based non-profit dedicated to education policy rooted in community vision.

Posted by: JusticeMatters | February 17, 2010 1:11 PM | Report abuse

So, it's common sense that linking teacher pay to test scores isn't good and solid academic research research backs it up and still Obama's US Dept of ED is pursuing this idiocy?

Thanks Valerie, for putting all this out there, but when oh when and how oh how is it going to get reversed before even more damage is done to the educational system?

Next I want to hear about the movement to convince Obama and Duncan that this is madness and must be stopped now.

Posted by: efavorite | February 17, 2010 4:41 PM | Report abuse

phoss1--you are full of the kool-aid that makes people equate testing with learning. You also lump Dewey together with Summerhill as examples of useless teaching philosophies and disparage the use of portfolios for assessments.

Maybe the next time you need a doctor you can request one who did great on multiple choice tests, but who has no hands-on clinical diagnostic experience.

Posted by: aed3 | February 17, 2010 4:51 PM | Report abuse

Bravo, Lisa Guisbond! - for pointing out that the arts and other subjects also have an impact on a student's educational life!

To reiterate: The simplist reason for NOT linking teacher pay to test results is that there are just too many variables (from the individual school environment to each student's background to the educational background of each teacher, etc.)to make it fair, worthwhile or meaningful.

Both Secretary Duncan and and President Obama need to be much more insightful and creative in addressing the whole picture of improving education.
Example: Instead of a salary scale type of incentive linked to test results, why not use "bonus" money to pay interested teachers to offer SMALL, EXTRA classes to underperforming students, or for doing EXTRA work in experimental methods that
can be tracked for interested parties, or for doing EXTRA coursework in their subject area beyond the required.....since teachers cannot be paid overtime because they are salaried, it is time to recognize that many good teachers are good because they put in EXTRA time, and that is a key area that deserves to be compensated.

Posted by: PLMichaelsArtist-at-Large | February 17, 2010 6:46 PM | Report abuse

Interesting idea, PLMicheals, to compensate teachers for extra work many of them are already doing - but if advanced student achievement is the only thing that counts, I don't think the Obama gang will go for it. Pains me to say it.

I'd like to think that if I could get in to talk to him for 15 minutes, I could talk him out of this ridiculous course, but I suppose there's something else going on that I don't know about -- something very bad that has allowed Obama to sell public education down the river.

I've never resented him sending his kids to private schools until now. They will be spared all this crap he's imposing all everyone who can't afford a private school.

If these ideas are so wonderful, why aren't private schools using them already? Why do people have to be bribed to use them?

Posted by: efavorite | February 17, 2010 8:16 PM | Report abuse

I say we should dump the salary and pay teachers like lawyers based on "billable hours". I'd easily double my salary if I got paid time and a half for overtime like the building service workers do. Please, stop treating me like a "professional". I'd make more if you treated me like a janitor!

Posted by: rsburton78 | February 17, 2010 9:45 PM | Report abuse

"If these ideas are so wonderful, why aren't private schools using them already? Why do people have to be bribed to use them?

Posted by: efavorite | February 17, 2010 8:16 PM"

Amen to that, I say, AMEN!

Posted by: edlharris | February 17, 2010 9:49 PM | Report abuse

Anti-testing folks should answer this:

What SHOULD teachers be evaluated on? If not student tests, then what?

I'm personally open to different ways of assessing students and teachers.....but let's get serious about proposing some alternatives rather than just carping on the sidelines..

Posted by: holzhaacker | February 17, 2010 10:22 PM | Report abuse

I think test scores should be part of a range of techniques for evaluating teachers, not the ONLY way.

But at the core of it all should be student progress.

The anti-accountability folks don't seem to think teachers should be held responsible for ANYTHING.

Sorry, but teachers need to be evaluated on what they're hired to do, which is TEACHING STUDENTS......and NOT on how many clubs they sponsor or whether they turn their lesson plans in on time.

Posted by: holzhaacker | February 17, 2010 10:25 PM | Report abuse

Teachers coul rated on student achievement... but a multiple choice test isn't a measure of student achievement. a proper assessment method, with portfolios and long-term observations would be way more expensive than anyone wants to spend.

Posted by: someguy100 | February 17, 2010 11:12 PM | Report abuse

I'm glad to see further attention paid to the problem of evaluating teachers based on test scores. Those who keep pushing this idea seem totally unconcerned by the lack of any research basis to support their ideas, and they suggest that any critic who neglects to mention the alternatives is "anti-accountability." Apparently bogus data is better than none at all.

Teachers can demonstrate effectiveness through National Board Certification, which requires evidence of student growth. Teachers can demonstrate effectiveness by assembling portfolios, participating in any number of robust and innovative evaluation programs that are available, and by meeting with peers and administrators to identify mutually agreed upon goals and measures of progress towards those goals. Student work, student surveys, and parent surveys are used in some places. The problem lies with school boards and administrators who do not invest in better evaluation, and to some extent with teachers who do not pursue it.

Meanwhile, this post doesn't even mention some of the most damning evidence against the testing=accountability crowd. Many tests are not instructionally sensitive, meaning that students' performance on the test cannot be tracked back to instruction - they can answer questions without any directly relevant instruction. That is a HUGE flaw in language arts/reading tests. Additional problems: sample sizes are too small - each student can represent too large a portion of overall results if results are grouped by teacher. Furthermore, students are not randomly placed, meaning that teachers do not teach equivalent classes. Teachers' assignments often change, along with changes in curriculum, resources, policies, and other school personnel, making year-to-year comparisons faulty. Research has also shown that "good" and "bad" rankings of teachers are not stable over a 1 to 4 year period - in other words, past results are not a reliable predictor of future success. Test designers and test experts say that it is invalid to use a test designed for one-purpose to carry out policies with another purpose. (e.g., see Braun, ETS). Add these problems to the unintended consequences of corruption and narrowed curriculum

Now, I would love to hear someone try to refute my position by offering the insights of their decades in the classroom and the studies they've reviewed. Where's the evidence?

Posted by: DavidBCohen | February 18, 2010 2:57 AM | Report abuse

Here’s a good YouTube spoof on the de-motivating effects of merit pay. It’s a hoot, really and too true, I’m afraid.

Duncan and Obama should see it. Rhee should see it too, but I’m sure it would have no effect on her. She has no discernible sense of humor and she already knows what’s right - it’s whatever pops into her mind.

Posted by: efavorite | February 18, 2010 8:20 AM | Report abuse

I haven't seen anyone mention a very important variable in assessing teachers'performance based on tests, and that is the PRIVATE TUTORING that many students receive via well-off parents or special programs (including special ed programs); the results of tutors' efforts are sprinkled throughout the various students' coursework.

How does one sift out how much is due to the primary teacher and how much is added (i.e. test points) from extra tutoring?

Posted by: PLMichaelsArtist-at-Large | February 18, 2010 9:48 AM | Report abuse

Indeed, PLMichael, it's not possible to sift out the effect of a tutor any more than it's possible to sift out the effect of an academically serious older sibling who helps his little brother with his homework, or a troubled older sibling who takes his little brother on drug runs.

There are simply too many variables affecting student learning to hold teachers completely responsible for them. Why can't smart people get this?

Posted by: efavorite | February 18, 2010 11:09 AM | Report abuse

To sum up the points of David B. Cohen above, when all is said and done and the kid sits down to take the test, the pernicious presence of "it's a crap shoot" hovers over the test taker. Hence, the inadequacy of over-correlating teachers' efforts to standardized test scores.

Posted by: Incidentally | February 18, 2010 5:57 PM | Report abuse

Where to begin? In "sound bite" land the concept of linking teacher pay to student performance seems perfectly reasonable. But, as with so many complicated issues, the devil is in the details.

First, we don't even have anything that resembles a national consensus on what "student performance" looks like. Is the high-stakes test given in Idaho comparable to the one given in Maryland or Massachusetts or Texas or Alabama? Are the tests even "graded" the same way? What is considered a passing score? That's certainly not consistent state-to-state.

What SHOULD U.S. students be able to do in 4th grade, 8th grade, 12th grade? What is an acceptable standard? What are the expectations?

Why are grammar and spelling not even taken into account when the infamous BCRs (Brief Constructed Responses) are graded on the Maryland Assessment? Are those skills unimportant? Is the person who can't punctuate or spell well an "educated" person? The test would have you believe so, but I couldn't disagree more Certainly something to consider when determining the "standard" we want our students to meet.

What if--as is the case at my elementary school--students leave the school after 3rd grade to go to a magnet program? Should my school (and the teachers who had those students K-3) be given any "credit" for the students' successes or does the magnet school get all the gravy? Something to consider when grading school performance.

What about students who have multiple teachers for reading or math? Who is "responsible" for that child's success or failure? Do the teachers split the "bonus" or share the "blame" if the kid does poorly?

Suppose a child takes the test a couple of days after his/her grandmother dies (as had happened in my class) or on the day his/her parents have a nasty fight, (as has happened in my class) or if the family is suddenly homeless (as has also happened in my class). Do those results reflect what the student knows? Should my salary hinge on this?

I understand that parents want to know that their kids are learning and I am willing to accept that I should be "graded" in an objective, relevant manner. But judging my teaching without taking into account the things I can't control and the vagaries of today's so-called "standards-based" learning is flat-out unacceptable.

Posted by: daveairozo | February 18, 2010 6:40 PM | Report abuse

There is no "anti-accountability" movement. Critics are saying that there are too many tests and that the tests are inappropriate. Nobody is saying eliminate all tests. The name of the group is FAIRTEST, not NOTEST.

Posted by: skrashen | February 19, 2010 2:49 AM | Report abuse

davairoso and others

These points you are making are very good. I really agree about the grammar and spelling on the BCRs. I understand the writing process completely (my students do well on BCRs) but have noticed that my own daughter cannot really spell. Spelling is not taught anymore I was told.
My point is that what is tested is what will be taught.

It is sad about the Georgia erasures. That simply teaches the kids to cheat. I saw something similiar in Houston when I taught there. (Years ago)

Posted by: celestun100 | February 26, 2010 4:18 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company