February was a cruel month for Governor Tim Pawlenty’s Quality Compensation for Teachers (Q Comp) program, Minnesota’s first attempt to couple teacher salaries to their students’ performance. It started with a front-page Star Tribune article, “Is it 'merit pay' if nearly all teachers get it?” — highlighting that only 27 of the 4,200 teachers enrolled in Q Comp failed to receive a raise during the 2007-8 school year.
Two days later, the Office of the Legislative Auditor released the first independent review of the program that found: “Q Comp’s effect on student achievement cannot be adequately measured using existing data.”
As the Governor attempts to implement Q Comp statewide, it is important to contrast his praise of the program with its increasingly well-documented shortcomings.
Over the last three years, roughly a dozen states have implemented performance-pay teacher reforms based on the findings of “value-added” educational research that quantifies the effect of an individual’s teacher’s instruction. Multiple longitudinal studies found that wide discrepancies in student learning — as much as a full year of additional instruction — can be tied directly to a pupil’s teacher, independent of class size or overall school quality effects.
From a fiscal standpoint, an excellent teacher costs as much as an ineffective one, meaning that improving teacher quality may be the most direct and cost-effective way of closing the racial achievement gap. Realizing these potential gains, however, will require significant change in not only how Minnesota pays its teachers, but how we recruit, train and retain them.
So, where does Tarvaris Jackson fit into the equation? Malcolm Gladwell’s timely review of the value-added movement, “Most Likely to Succeed,” compares the process of identifying and training effective teachers to successfully drafting a NFL quarterback.
Every year a new crop of highly touted college quarterbacks is selected early in the NFL draft. From a franchise’s perspective, the difference between drafting the right or wrong prospect will not only be measured in wins and losses, but in millions of dollars of revenue. One would assume that, over time, this pressure, plus all the upfront scrutiny and the opportunity to evaluate the results, would create a correlation between when a quarterback is selected in the draft and his performance in the NFL. It doesn’t. Superstars emerge both from the draft’s upper and lower echelons.
Peyton Manning, three-time league Most Valuable Player, was drafted first in 1998, but other prospects like the Patriot’s Tom Brady — drafted 199th overall, as a fourth-string back-up — can become equally successful. And for every unrecognized late-bloomer there are expensive, high-profile flameouts. Remember this guy?
The quarterback problem is a reflection of the fact that successfully adapting to the NFL is not just a question of physical ability. Rather, it requires development of a complex blend of physical and mental attributes that can be learned, but necessarily taught. Peyton Manning was able to make those intangible adjustments, where other gifted athletes — the Vikings' Tarvaris Jackson, for example — struggle to do so.
Teachers face a similar do-or-die transition during their first several years in the classroom. Some develop the ability to make split-second decisions and assessments that create a highly-functioning classroom. Others flounder, year after year, trying to cover a full year’s worth of curriculum. Until the advent of the value-added movement, however, no one was tracking in quantifiable terms who succeeded or failed.
The performance of quarterbacks is tracked and analyzed in multiple dimensions, play by play, game by game and season by season. Measuring the effectiveness of teachers is much more difficult. The only comparable measure of a teacher’s quality of instruction is their students’ standardized test scores, an extremely small sample size.
In order to isolate the effects of a teacher’s instruction, it takes a minimum of two, preferably three, years of testing data. Even then, only the top and bottom 15% of teachers can be separated from the mean. Given these statistical realities, any approach to true teacher quality reform must do several things: reward the most effective teachers, remove under-performing teachers and create a data infrastructure that can assess a teacher’s effect on her students.
Q Comp fails to even accomplish the one thing we could reasonably expect it to — raise the salaries of high-quality teachers.
In its current form, value-added analysis is only one tool among several that a school district can use to create its own definition of teacher merit. But when “merit” is defined along traditional lines, i.e., time-of-service and professional credentials — neither of which has been strongly correlated with increased student achievement — we should expect the status quo to continue.
If the Governor wants Q Comp to deliver its advertised reforms, he must mandate the use of quantified measures of teacher merit. Otherwise, the majority of districts will follow the path of least resistance and minimal results, all at increased taxpayer expense.
Minnesotans, I believe, would prefer the use of over-hyped, statistically dubious strategies be limited to the Metrodome, and kept out of their childrens' classrooms.
— Kevin McNellis is an intern at Growth & Justice