Cross-posted from Education Week
Over the last two weeks, I shared with you a model state plan for education accountability. In this blog, I explain why I chose its main features.
It starts with four state tests given to all students during the course of their careers in public school. The first is a diagnostic test of entering first graders, designed to test their readiness for first grade. Young people enter the first grade with very different degrees of readiness for first grade, even if their native language is English. Their chances of success in elementary school and thereafter are greatly affected by the degree to which the education they receive in the primary grades is geared to their starting point. Strictly speaking, these tests are not an accountability measure, but would be a crucial feature of the testing regime and could well make the difference between a successful and an unsuccessful experience in the primary grades.
Then I propose three more tests of every student during their career in the public schools, the first at the end of the primary grades, when they should have mastered the essentials of reading, the second at the end of middle school and third in high school. All of them would be designed to capture as much as possible of the cognitive and non-cognitive aspects of the kind of education now widely believed to be needed for a student to go on to a successful life, far more than is now captured in the typical state test. While it is possible to do this, it is expensive. But, by testing all students less often then we now do, we could get much better tests that we could afford. Because what gets tested is what is taught, this is the only way that we can prevent our accountability testing system from narrowing what gets taught to what can be cheaply tested. The plan would embrace the subjects implicated in the idea of what it means to be an educated person. This would prevent the accountability testing system from driving out of the curriculum subjects that almost everyone agrees are very important.
The elementary and middle school tests would be designed for specific grades. High school would be different. The high school test would be designed to assess the degree to which students had mastered the skills and knowledge needed to be successful in the first year of a typical first year community college program, which is the minimum standard for being successful in both work and college, since most vocational education in the United States is offered in our community colleges and students can transfer to four-year state colleges after two years of the appropriate academic program in community college. High schools would be measured by the degree to which they were able to get all or almost all of their students to that standard. This would be a performance standard, not a time-in-the-seat standard. Some would achieve it as early as the end of their sophomore year, others not until the end of their senior year, but all would achieve it. As soon as they did so, they could go immediately to a community college or stay in high school and take a rigorous college-prep program designed to get them into a selective four-year college. American high schools would cease to be sorting institutions; it would be the end of tracking. All students would have a route to two-year and four-year college programs—whether academic or vocational—and all would leave high school ready to succeed in those programs. Not least important, this new accountability system would be designed to hold students, not just teachers, accountable. All students would have a strong incentive to take tough courses and study hard in order to achieve their dream. One has only to look at high schools in top-performing countries to see how powerful this idea can be.
The state would plan backwards from the demands of its final high school courses to create a curriculum framework spelling out the progression of topics and competencies students could be expected to study, grade-by-grade, based on what is now known about how students actually develop through the years, subject-by-subject.
In between the tests of all students would be tests of only mathematics and English competency in the off-years. The student population of each school would be sampled. Minority and low-income students would be oversampled to make sure their performance was accurately portrayed for each school. Obviously, the costs of this off-year testing system could be cut dramatically by testing every second year instead of every year. In any case, these tests would be used to help state authorities identify schools that needed a closer look from a team of experts, who would pay an extended visit to the school to understand why it was underperforming and help the school and its community to develop a plan for turning the school around.
That plan would very likely include a strategy for strengthening school performance by pairing the school with a high-performing school, transferring some of the staff to a high-performing school, transferring some staff from a high-performing school to the school in need or a similar strategy. Large districts would be expected to make these arrangements using their own teaching resources. The state would make them for smaller districts. In all cases, the state would recognize an obligation to provide more and better teachers to schools serving the hardest-to-educate students than the easier-to-educate students.
One of the most powerful changes to the prevailing accountability system in my proposed plan is the introduction of the kind of accountability most common in the high status professions but largely absent in school teaching—accountability to one’s professional colleagues. Adoption of real career ladders in our schools, combined with a regime in which teachers would be expected to work closely with each other—and have the time to do so—would change the culture of the school. The performance of each teacher would be visible to all and the reputation of the school would depend on the actions of all the teachers, creating a large incentive for the best teachers to deal with the weakest teachers. This is the best accountability system of all.
Under this plan, the state would be obligated to routinely make public a wide range of data related to the performance of every public school in the state on a web site easy for parents and other non-specialists to use. The system would provide detailed data on the performance of minority and low-income students at each school, as well as the performance of the school as a whole. That data would be used to frame the state’s actions designed to improve school performance, whether for vulnerable students or the student body as a whole.
But the system would not be designed to try to attribute the performance of the students to particular teachers or to produce data that would be used to determine whether individual teachers were retained or let go. As I have explained elsewhere, the weight of evidence shows that such systems not only do not work; they actually accelerate the rate at which good teachers leave teaching and deter good candidates from entering the profession. Not least important, such systems require annual census testing of all students, which greatly increases the amount of testing required and thereby produces strong incentives for states to narrow testing only to mathematics and English literacy and to use cheap tests which cannot measure much of that which is important for students to know and be able to do.
The accountability regime I have proposed will not itself produce results comparable to those the top-performing nations enjoy. For that to happen, the United States will have to provide more resources for schools serving the hardest-to-educate students than for schools serving the easiest-to-educate students. The country will have to recruit its teachers from the top quarter of high school graduates and provide them with a far more rigorous professional education than they have received in the past. Schools will have to do a much better job of providing an environment in which teachers are supported to continuously improve their expertise. Much more attention will have to be given to the development of powerful instructional systems to match the new standards, curriculum frameworks and assessments. But I have no doubt that the kind of accountability system I have proposed will work far better than the one described by No Child Left Behind as modified by the Race to the Top program and the waivers granted by the U.S. Department of Education.
