After the punishing No Child Left Behind years – years when testing was used to set impossible goals for schools and teachers and was an obstacle to real teaching – you now hear a lot about rejecting standardized testing altogether. But Andy Hargreaves and Henry Braun, professors of education at Boston College, take a more balanced approach.
In their recent policy brief called “Data-Driven Improvement and Accountability,” published by the National Education Policy Center at the University of Colorado, they show how test results, used intelligently together with other data, can be used to help schools improve, not just beat them up. They say the way we’ve been doing it is all wrong and they make recommendations for a better way.
Listen to them talk about it here on American Radio Works (or download it as a podcast):
The paper itself is written with academic precision, but don’t give up. It’s beautiful. They talk more sense than you’ve ever heard before about testing.
I have summarized some key highlights below. The numbers refer to bookmarks I’ve added to the pdf version of the paper, so you can see in context what they actually said. (If you don’t see the bookmarks in your browser, save the pdf locally and open it in your Adobe reader. Or just skip down the red highlight rectangles in order.)
Here’s the gist:
- If we want to use tests to improve schools, we need to work together to make the data accurate, meaningful, fair, broad and balanced. When bosses use test results to punish, school don’ t improve. (1)
- When we use tests merely to bash teachers and schools, rather than to help improve them, we damage the schools. (2)
- The winning countries and companies and sports teams use tests for improvement rather than accountability.(3)
- Over the past 10 years in which top-down, high stakes testing has been used in the U.S., it hasn’t worked. (4 & 5)
- But there are good systems for testing and making use of the test results. In Ontario, for instance, test results lead to improvement efforts, but not to ranking or grading the schools or teachers or to closing schools. And test scores are used together with other measures to figure out what needs to be done [Note: This is very much like the NH approach]. (6)
- There is wide agreement that Ontario is one of the best illustrations of how to use data for both improvement AND accountability. There is pressure for improvement, but it is not based on threats: the assumption is that poor performance comes from a lack of capacity rather than a lack of effort. In the U.S., most states use their accountability systems to focus strictly on outcomes rather than as a tool for actually working on improving the schools….When U.S. schools don’t improve, there may just be too many obstacles. Children in the most disadvantaged communities often find themselves not only facing disadvantage and instability at home, but also experiencing an unstable and under-qualified teaching force, high principal turnover, and a politically disruptive environment of constant change, repetitive reforms and school closures that further magnify the insecurities in their lives in school. In these cases, more test data won’t help. (7)
Hargreaves and Braun make a dozen recommendations (starting at bookmark 8). They say things like: If we want critical reasoning, emotional and social learning, creativity and teamwork, then we should measure that, not just test for what’s easy to test for. “Test prudently, not profligately,” meaning things like testing fewer grades, using a sample of students rather that every student and rotating subjects year-by-year. Set challenging standards. Establish a collaborative atmosphere of high trust among teachers and administrators to make use of the data. And a lot more.
In their conclusion, they say that debates about using data will not be about whether to use the data. The information should not be ignored. The question is how to benefit from the data without falling into the traps. The essential choice is “…whether it will be used to enhance and enrich the quality of collective professional judgment so that all America’s students will reap the benefits of a better education.”
Clearly, recommendations like reduced testing would require federal legislation but there is a big agenda here even within what states and school districts can do on their own.
NEPC has published, along with the policy brief, model state legislation prepared by Kathleen Gebhardt of Children’s Voices. It is, essentially, the legislative version of the Hargreaves/Braun policy brief. Here’s my bookmarked copy. The key concepts include:
- Accountability is broadly defined to include everyone from state policy makers to teachers. (bookmark 1)
- Assessment is also defined broadly to include all subject areas and academic as well as non-academic standards set by the state board of education.
- It introduces the concept of “capstone” projects that show “mastery of essential skills and learning” and “should demonstrate a student’s ability to think critically and creatively, to solve problems, and to communicate effectively.” (3)
- The bill defines audits and evaluations and the teams needed to do them. (4)
- The bill mandates that classroom teachers develop classroom level formative (day-to-day), summative (periodic snapshots of overall progress) and capstone assessments.(5)
- At the district or school level, the goal is to use existing data in newly informative reports and then goes on to list 29 data items to be reported on. (6)
- At the state level, the bill sets out the quality standards the statewide assessment must meet (7) and goes on to mandate a sampling strategy that would dramatically reduce the amount of testing to which a student was exposed.(8)
- Then the bill sets up a system of state inspections, saying that, “The system is designed to move the State away from producing assessment results that can or will be used for high-stakes evaluative decision-making or that are likely to prompt instruction geared toward test-preparation.” (9) The bill goes on the define how the inspection system would work and what it’s goals would be. The inspection teams would use the locally developed reports and well as develop a range of data of their own – 19 specific items are listed (10). They would inspect the schools that needed the most help, but get to every school at least every four years.
- Schools must develop improvement plans, defined here in some detail, in response to the inspection team reports (11)
- Every five years, there would be an audit and evaluation of the whole system by the state auditor’s office. (12)
- The final section defines the grievance process available to just about anyone. (13)
Much of this is beyond the capacity of the state of New Hampshire, at least as we conceive of the state role in education today. However, as our testing strategies evolve, there’s a lot of food for thought here.