« New Jersey Advocacy Groups Didn't Get Their Stimulus Way | Main | South Carolina Gov. Mark Sanford: Unleashed »

Duncan to States: Test Scores and Teacher Evaluations Do Mix


Schools and the Stimulus
In the first in a planned series of four speeches about the Education Department's top reform priorities, Secretary Arne Duncan told education researchers today that innovation and new practices must be supported by evidence-based research. And, what's more, he took states to task for enacting laws barring student test scores from being used in teacher-evaluation decisions.

Debbie Viadero, who covers and blogs about research for EdWeek, called in to say that Duncan, in particular, singled out New York and California for having such state laws.

But the problem goes deeper than those two states. According to the latest update from the Data Quality Campaign, 17 states have no plans to create a unique teacher-identifier number and link that number to student achievement data. My colleague Stephen Sawchuk, who blogs over at the Teacher Beat, has explored this issue in depth.

In his speech to the Institute for Education Sciences, Duncan urged researchers to work on improving accountability models based on student achievement growth on test scores and developing fair models of compensating teachers and other school staff based on the achievement of their students. Ultimately, he added, the data should be used to ensure that students are on track to graduate and to succeed in college, according to an Education Department press release.

Duncan's planned series of speeches is built around the four "assurances" that are included in the federal economic stimulus law. The assurances say that states must make progress in turning around low-performing schools, building better longitudinal data systems, recruiting and retaining effective teachers, and implementing more rigorous college- and career-ready standards. States must show such progress, or a plan for progress, to get their share of the $48.6 billion State Fiscal Stabilization Fund. In addition, Duncan has said he will use a state's progress on the assurances to help him make decisions on who gets the billions of dollars in Race to the Top funding under his control.

This isn't the first time Duncan's put a little bully in his bully pulpit. He's put states on notice that charter school caps that limit the growth of these nontraditional public schools are not his idea of innovation or reform.


"Innovation and new practices must be supported by evidence-based research."

But there's no evidence-based research to support his emphasis on teacher performance pay or charter schools.

Oh well.

I believe targeting teachers to make major improvements in the education of our American children is off the mark. It's no secret that test scores highly influenced by socioeconomic environments. China and other "well educated" students are indoctrinated with the importance of education by parents and society early in life. Our challenge is to improve academics performance without destroying individualism and capitalism. Social Emotional learning is needed to help our American children adapt to many cultural and diverse family changes brought on by much freedom without healthy competition and a non-intrusive amount of responsibility-loyalty without extremism. We miss the forest for the trees. Education was desirable in 1950-1960's so we have to find out what has changed since then?

Test scores can validly provide an indication of a problem (students are not learning), but don't provide information related to the causes (socio-economic, poor materials, dangerous facilities, or bad teaching).

Each one of these needs to be examined to determine their contribution to the problem. And there is a way to provide objective feedback to all stakeholders regarding the teaching and resulting student behaviors within the classroom. New technology tools such as the eCOVE Observation Software (http://www.ecove.net) can be used to gather objective data on teacher and student behaviors. Match that data to solid research on best practices and you have the basis for solid reflection and decision making. We need to abandon the checklist and observer opinion/judgments and use real data; now that it's easy to collect there's no excuse.

The other areas of influence need to be examined with the same objective basis. It's a complex problem with a complex solution, and too important to 'solve' by finger-pointing.

Comments are now closed for this post.

Follow This Blog


Most Viewed on Education Week



Recent Comments