One of the wisest and sanest voices in the nation on the subject of teacher quality, teaching quality and teacher evaluation is Linda Darling-Hammond of Stanford University. Linda has been involved for many years in studying these issues and working directly with teachers to improve practice. During the presidential campaign of 2008, she was Barack Obama’s spokesman and chief adviser on education, but was elbowed aside by supporters of Arne Duncan when the campaign ended. The Wall Street hedge fund managers who call themselves Democrats for Education Reform (they use the term “Democrats” to disguise the reactionary quality of their goals) recommended Duncan to the newly elected president, and you know who emerged on top.
Linda, being the diligent scholar that she is, continued her work and continued to write thoughtful studies about how to improve teaching.
After the 2008 election, the issue that predominated all public discussion was how to evaluate teachers. This was no accident. Consider that in the fall of 2008, the Gates Foundation revealed its decision to drop its program of breaking up large high schools. Recall that the foundation had invested $2 billion in breaking up big schools into small schools, had persuaded some 2,500 high schools to do so, and then its researchers told the foundation that the students in the small high schools were not getting any better test scores than those in the large high schools.
Gates needed another big idea. He decided that teacher quality was the big idea. So he invested hundreds of millions of dollars in a tiny number of districts to learn how to evaluate teachers, including thousands of hours of videotapes. Where Gates went, Arne Duncan followed. The new Obama administration put teacher quality at the center of the $5 billion Race to the Top. If states wanted to be eligible for the money, they had to agree to judge teachers–to some considerable degree–by the test scores of their students. That is, they had to use value-added assessment, a still unformed methodology, in evaluating teachers.
In response to Race to the Top and Arne (“What’s there to hide?”) Duncan’s advocacy, many states have now passed laws–some extreme and punitive–directly tying teachers’ tenure, pay, and longevity to test scores.
No other nation in the world is doing this, at least none that I know of.
The unions have negotiated to reduce the impact of value-added systems but have not directly confronted their legitimacy.
After much study and deliberation, Linda Darling-Hammond decided that value-added did not work and would not work, and would ultimately say more about who was being taught than about the quality of the teacher.
The briefest summary of her work appears in an article in Education Week here.
She recently published a full research report. Here is a capsule summary of her team’s findings about the limitations of value-added assessment:
“Measuring Student Learning
There is agreement that new teacher evaluation systems should look at teaching in light of student learning. One currently popular approach is to incorporate teacher ratings from value-added models (VAM) that use statistical methods to examine changes in student test scores over time. Unfortunately, researchers have found that:
1. Value-Added Models of Teacher Effectiveness Are Highly Unstable:
Teachers’ ratings differ substantially from class to class and from year to year, as well as from one test to the next.
2. Teachers’ Value-Added Ratings Are Significantly Affected by Differences in the Students Assigned to Them: Even when models try to control for prior achievement and student demographic variables, teachers are ad- vantaged or disadvantaged based on the students they teach. In particular, teachers with large numbers of new English learners and students with special needs have been found to show lower gains than the same teachers when they are teaching other students. Students who teach low-income stu- dents are disadvantaged by the summer learning loss their children experi- ence between spring-to-spring tests.
3. Value-Added Ratings Cannot Disentangle the Many Influences on Student Progress: –––Many other home, school, and student factors influence student learning gains, and these matter more than the individual teacher in explaining changes in scores.”
The application of misleading, inaccurate and unstable measures serves mainly to demoralize teachers. Many excellent teachers will leave the profession in frustration. There will be churn, as teachers come and go, some mislabeled, some just disgusted by the utter lack of professionalism of these methods.
The tabloids will yelp and howl as they seek the raw data to publish and humiliate teachers. Even those rated at the top (knowing that next year they might be at the bottom) will feel humiliated to see their scores in the paper and online.
This is no way to improve education.