Our educational system is waltzing with a crowd of reformers who have the hubris to think they know a cure to every ill - and the cure always has something to do with test scores. The latest fix focuses on teacher quality, and proposes to improve evaluation by “unleashing the power of data,” as Secretary Duncan put it. But a new report was released last week that suggests this data may be unreliable for the jobs it is being asked to do.
Some of the nation’s leading education researchers, including Diane Ravitch, Linda Darling-Hammond, Eva Baker and Richard Rothstein, co-authored this report, entitled “Problems with the Use of Student Test Scores to Evaluate Teachers,” and available for download here.
Here are some of the issues the authors highlight:
There are many factors other than teachers that have a significant effect on student outcomes. These include:
• Other teachers the student has had
• School attendance
• Learning at home and in the community
• Family support
• Mobility
• The effect of peers
• Summer learning loss - especially profound among low-income students
All of this means that when value added methods are applied to teachers of low-income and English-learning students, the teachers have a harder time achieving the expected growth.
Going beyond the limitations of the data, the authors raise concerns about the effect of focusing teacher evaluations on test scores. They write:
Research shows that an excessive focus on basic math and reading scores can lead to narrowing and over-simplifying the curriculum to only the subjects and formats that are tested, reducing the attention to science, history, the arts, civics, and foreign language, as well as to writing, research, and more complex problem-solving tasks.
They also point out that this trend will discourage teachers from choosing to work in schools with the neediest students. Here in Oakland we already see evidence of this, as turnover rates are the highest in the lowest-performing schools. Systems that reward teachers for their individual growth also can discourage collaboration by creating a competitive environment within the school.
When we think about the whole thrust of this effort, it is aimed at creating “accountability” for teachers. We can agree that we ought to be accountable for the quality of our work, but we must build that accountability on a solid foundation of shared values. If I were to evaluate a restaurant merely on the popularity of food it sold, Dunkin Donuts might emerge as the greatest in the land! Every serious consideration of this issue must return to this point.
Many of the values we actually seek to elevate in our schools - honesty, creativity, initiative, the ability to apply knowledge in solving problems - these things are not measured well by the tests we are using. We need systems of evaluation that are capable of encompassing the quality of teaching relative to these dimensions.
This week we heard that even in China, where the educational system has prized rote learning for millennia, Premier Wen Jiabao gave a lengthy speech calling for change. He said:
Students don't only need knowledge; they have to learn how to act, to use their brains. As Einstein said, imagination is more powerful that knowledge.
We must encourage students to think independently, freely express themselves, get them to believe in themselves, protect and stimulate their imagination and creativity.
If it is good enough for a billion Chinese people, it ought to be good for us.
Update: For a chuckle -- and a very serious point -- take a look at School Finance 101’s post on the value-added issue. The author riffs on a satirical “news” clip from The Onion, which observes that students who do not care, tend not to score so well. And perhaps these students are not randomly distributed! Doh!
What do you think? Should we “unleash the power of data” for the purpose of evaluating teachers? Or should we keep data on a short leash?