Some people believe using student data to evaluate teachers is unfair:
“In the field test given a year ago, 91 percent of English Language Learners and nearly 80 percent of low-income students did not meet proficient. My class is comprised of 40 percent English language learners and nearly 100 percent are low-income. Because new state legislation … will link my students results to my professional evaluation, this does not bode well for me or for my colleagues.” – Emily Talmage, teacher (link)
“Test scores being used to evaluate teacher performance are problematic from the get-go. If that is the way we choose to evaluate teacher performance, it is certain that results will be skewed in favor of school districts that are in the higher socio-economics category, demographically speaking.” – John Goodman, former teacher, Oswego mayor, and co-chair of the New York State Democratic Party (link)
But test data can be used in many ways – not all of them unfair. In value-added models, teachers are evaluated based on how much students learn – aka their growth – during a school year, not on their absolute levels of achievement.
Achievement-based evaluations, therefore, ask if students are performing exactly at a level that puts them on track to graduate college and career ready. It does not look at where students started.
Value Added evaluations, on the other hand ask if achievement scores are where we expect given students’ past performance. It’s based on students’ prior achievement history and assesses student progress relative to similar students. It accounts for where students start, and measures the impact teachers have on student learning.
Value Added generally works like this (see Hechinger Report, Harvard’s Strategic Data Project, and Corcoran and Goldhaber for more details):
Typically, students will score within two standard deviations of what was predicted 95% of the time. The key is identifying the 5% of teachers who, over the course of multiple years, have students scoring over two standard deviations above or over two standard deviations below predicted.
So yes, achievement scores will be lower when assessments are more rigorous. But, it usually doesn’t matter for value added, because value added is a relative measure.
The opt-out movement is being egged on by teacher unions in hopes of disrupting value added because they don’t want teachers to be evaluated on the basis of standardized tests.
They – justly – point to reputable studies that caution that error rates are too high to use Value Added in personnel decisions, that small programming errors in formulas result in teachers assigned different ratings, and that school variables need considered. These are valid concerns and statistical models need to assure fairness. Multiple years of data need to be used and interventions need to be tried before any negative personnel action is taken.
But, in no state does Value Added comprise more than 50% of a teacher’s evaluation, and fairness to students is important too.
One study, for example, has shown the positive relationships between effective teaching (as measured by their value-added impact estimate) to long-term student outcomes. Students assigned to high-value-added teachers are more likely to attend college, earn higher salaries, live in wealthier neighborhoods, and save more for retirement, and are less likely to have children as teenagers.
Additional studies have also shown that poor and minority kids are much more likely to have less effective teachers (see here and here for examples). In New York City, for example, black and Hispanic students and high poverty students are more than two times as likely to have a teacher rated “ineffective” or “developing” as white or low-poverty students.
It is precisely the students with learning challenges that Ms. Talmage and Mr. Goodman reference who need objective, annual measures of progress. It is precisely their teachers that benefit from Value Added because it goes beyond a simple one-time snapshot of achievement and illuminates teachers that are realizing remarkable growth with students who haven’t yet reached proficiency.
Providing our most vulnerable students with equal opportunity means, more than anything, assuring equal access to quality teachers. Value Added, complemented with fair observations and quality feedback, is part of the means to get there.