Insight
Improving teaching: the real value of value-added
Apr 10, 2008 10:21 AM
“There is a way [teachers] can see how they did last year with their low-, average- and high-end kids. Then ask, ‘What can I do in my classroom to stretch the instruction and make really good progress all across the board?’”
— William Sanders, pioneer in value-added measurement
Most teachers would agree with the research showing they are a key factor — probably the key school-based factor — in student achievement growth. But they don’t feel so valued when that research starts to get turned against them.
Yes, no kidding, teaching can make a huge difference. But now dozens of the country’s largest school districts have seized on using simply numbers through something they call value-added measurement. This is a statistically complex way of purportedly calculating teacher contribution to student growth, as a way to judge teachers — weed out bad ones, promote good ones, award merit pay and assign blame.
In doing so, they have strayed far from what William Sanders — a pioneer in value-added measurement — found most useful about his system: it was a way to help teachers improve their teaching and, by that, raise student achievement.
Let teachers own the data
“To me the real value of value-added assessment is for teachers to have for themselves some way of checking on how they are doing,” Sanders said in a recent interview. He fought successfully to prevent teacher-level data from being released to the public in his home state of Tennessee, and “18 years later I’d be just as strong an advocate of confidentiality as I was then,” he said. In Tennessee, the teachers own their value-added data and can share it or not, completely voluntarily.
Sanders doesn’t play “gotcha” with his assessment system, which measures the difference between a student’s expected achievement and actual achievement. Rather, he wants “to get teachers and principals to think about data in an entirely different way,” he said. “I think of value-added as the beginning block of formative assessment. There is a way they can see how they did last year with their low-, average- and high-end kids. Then ask, ‘What can I do in my classroom to stretch the instruction and make really good progress all across the board?’”
Sketchy versions
In the current climate, Sanders is a minority voice. Instead, value-added is too often being done to teachers instead of with them. Policy-makers or administrators are rushing to implement the sketchiest of value-added systems for the most malicious purposes.
In an April 2006 discussion paper, Robert Gordon, Thomas Kane and Douglas Staiger of the Brookings Institution called for states to measure teacher effectiveness by student performance over two years and establish “a presumption” that the bottom quarter of teachers, measured in this way, should not qualify for tenure and be barred from teaching.
In Florida, after skipping over the fine print on value-added, legislators and state administrators rushed a merit pay program into place in 2006 that ranked teachers in a school based mainly on student test score gains. Now referred to as the “Hillsborough fiasco,” the program, as first implemented in Hillsborough, Fla., wound up awarding bonuses mainly to teachers in the wealthiest schools. Sixty out of 67 Florida school districts walked away from the money the following year.
In the face of such wrong-headedness, it may be that it’s up to teachers to make intelligent and informed use of value-added. “Value-added estimations can be a powerful tool that helps teachers design and improve their instruction if (and only if) the results are seen as supportive and not punitive,” says Rob Weil, AFT’s deputy director of educational issues. “It’s unfortunate that some districts are not designing their systems in ways that are beneficial to kids.”
Looking at growing kids
Many educators believe value-added measurements hold real promise for schools. It makes sense, if you have the data, to look at student growth over time, try to figure out which students progressed most rapidly and try to adjust instruction to capture the conditions that seemed to work. Given the data, who wouldn’t explore it?
“These data tell you how well you grew a group of kids in a particular subject area,” said Katie Hartley, a 5th-grade teacher from Ohio whose school has used value-added assessment for several years. “For example, if you had high growth in math but not in reading, you ask yourself, what did I do differently with math, what kind of materials did I use?”
There are no consequences attached to the reports in Hartley’s school. No one is paid more or less, hired or dismissed, based on the assessments. But the teachers are given time to write action plans based on the data.
“Under the current system where we put so much emphasis on passing or failing, the kids at the extremes get lost — the really high and really low ones. When you look at value-added data it forces you to look at how you did at growing those kids,” Hartley added.
Rewrite the plan
To make it work, educators and data specialists must rewrite the goals of value-added. If the goal is to brand teachers using student test scores, then the likely consequence will be narrowing the curriculum and focusing on the test. But if the goal is to compare two reading programs, for example, to see which one worked best; to find ways to boost achievement for students who are in the middle-ability range in a classroom; or identify students who are not working to their potential, then value-added data can give teachers evidence, ideas and direction.
For another, educators must steer the conversation away from performance outcomes to focus on instructional inputs. “We should observe what a teacher is doing to get added value,” advised one educator at a national conference on performance incentives in Nashville, Tenn., in February, where value-added was by far the biggest topic.
At the conference, top researchers were skeptical about the use of value-added to evaluate teachers.
“Getting it right is going to be exceedingly difficult if not impossible,” said Harvard researcher Paul Peterson.
“We do not get anywhere if we substitute one narrow quantitative-based system for another” in evaluating teachers, warned author Richard Rothstein.
Pay-for-performance measurement tools are “too formulaic — they take the creativity out of the profession,” complained National Council on Teacher Quality’s Kate Walsh.
But flying under their radar, value-added is already being put to use in classrooms and schools to inform instruction. It seems to work where the aim is not to blame or sort teachers, or pay them less or more, but to coax the greatest achievement from students.
