You may remember that I moonlight as Education Week ‘s assessment reporter in addition to covering teacher issues. Right now, I’m in Boston covering the U.S. Department of Education’s first public forum on the $350 million that it’ll be putting toward consortia of states that create common assessments aligned to common reading and math standards.
The panelists are having very rich conversations on everything from how to use technology to improve what cognitive skills can be measured to how to structure consortia that work together effectively (do you need an executive director?). Yesterday, though, much of the conversation focused on the role of teachers in the assessment process.
A number of the panelists brought up the idea of using performance-based tasks and extended open-response questions, and engaging teachers in the design, use, and even the scoring of these tasks. A Canadian testing official, Jim Dueck, talked about how the province of Alberta has done this for a number of years. To minimize bias and scoring snafus, teachers read all of the responses blind, and the responses pass through many scorers and review levels.
The benefit, proponents say, is that teachers then have much more understanding of the standard at hand and whether students truly demonstrated mastery of that standard. They also feel more ownership of the assessment process, rather than facing decontextualized assessements dropped out of the sky.
The panelists didn’t agree, though, on whether such tasks should be performed in a low-stakes context in the classroom, to guide instruction, or whether such tasks could be rolled into an accountability context in a fair, reliable way.
The only state that did anything like this under the No Child Left Behind Act, Nebraska, faced a lot of skepticism about the system because its assessments were locally designed, making it harder to develop comparability across districts. And in 2007, when House education leaders proposed allowing 15 states to develop a similar system for NCLB, I remember that Dianne Piche, then at the Citizens’ Commission on Civil Rights (now at the Education Department), likened permitting teachers’ scoring of their students’ tests to allowing 16-year-olds to score their own driver’s-license exams.
But could it work if the tasks were standardized across a consortium? What if scoring were done blindly, as in Alberta? How about some kind of centralized auditing process of teacher scoring to increase reliability and comparability?
Do the challenges outweigh the costs, or vice versa?
Tell us what you think.