Remember when the "value added" to a child's life by a teacher included more than raising test scores--when improvement in standardized test data was a by-product of a solid year of learning, rather than the goal? Remember when the goal was uncovering, then capitalizing on, an individual child's capacities and talents, building their useful knowledge base, sharpening their skills?
Oh, wait. We don't live in that soft, humanistic--dare I say progressive?--world any more, the world of the whole child, whose learning progress need not follow a frequently-measured linear growth trajectory to be deemed satisfactory. We are now in the era of "scientifically based" evaluation.
Great new piece out on Value Added Methodology from EPI. Turns out that VAM, in its present form of development, resembles what the What Works Clearinghouse fans sneer at as "junk science"--or, at the very least, is not a reliable, stand-alone metric for determining teacher effectiveness. Lots of great, educator-friendly information on the report's findings here and here and here.
One thought and two scenarios on VAM:
First, whoever named this statistical modeling protocol "value added" (Bill Sanders?)--and tied it to teacher "effectiveness" was a genius. Because who wouldn't want to add value to a child's education and life prospects? When you understand the clear, working definition of the numeric value being appended, however, the noble-sounding descriptor "value" feels a lot less...valuable.
Scenario One: A graduate seminar in education policy at a well-respected university. Famous Economics Professor is presenting an innovative statistical model for determining teachers' ability to
raise test scores improve learning. A pointed question is asked about the structure of the equation: Shouldn't this sigma go there, instead? Shouldn't the value of T be squared, when the value of X is less than R divided by L? Around the table, eyes glaze over, although nobody wants to admit that they're befuddled. An argument breaks out--four Asian psychometricians do battle with Econ Professor over the usefulness and validity of the statistical model. None of the grad students (who are preparing for high-profile careers in education policy) says anything.
Scenario Two: A statewide task force on identifying and formally acknowledging teacher quality. High-ranking ed department official says that in the state's Round Two RTTT application, the state has identified "effective" teachers as those who can produce a year's worth of value-added growth. A "highly effective" teacher will produce growth equivalent to a year and a half. She notes anything less than a year and half will not impress those evaluating the RTTT applications.
Superintendent of a high-performing suburban school:
Say one of my best third grade teachers gets a child smack in the middle of the third-grade achievement spectrum. In order for her to be considered "highly effective," the teacher must take the child not just to fourth grade but a half-year beyond? And the fourth grade teacher who gets that same child must move him another year and a half, to sixth-grade level? And by the time the child actually reaches sixth grade, he must be performing at the ninth grade level, for all his teachers to be considered highly effective? By the end of 8th grade, the child who's had only highly effective teachers should be ready to graduate. Right?
Is that even possible, unless the child has somehow accessed and mastered the 7th and 8th grade curriculum--and the entire high school course sequence? Isn't there eventually a ceiling effect, where growth slows, due to the maturity of the child--even a very bright child, taught by the best teachers?
Isn't it easier for a teacher to show a year and half of growth if the child coming in to her classroom is a year or two behind? And what's the long-term value in having a child show growth several years ahead of his chronological age? Does that correlate with success in college?
Good questions. The high-ranking official's answer: that's what we've been told the Department is looking for--the strategy that will give us more points in the Race to the Top.
In other words, added value.