Fixing Our Broken System Of Testing And Accountability: The Reauthorization Of ESEA
** Reprinted here in the Washington Post
Our guest author today is Stephen Lazar, a founding teacher at Harvest Collegiate High School in New York City, where he teaches Social Studies. A National Board certified teacher, he blogs at Outside the Cave. Stephen is also one of the organizers of Insightful Social Studies, a grass roots campaign of teachers to reform the newly proposed New York State Social Studies standards. The following is Steve’s testimony this morning in front of the Senate HELP committee’s hearing on ESEA reauthorization.
Sen. Lamar Alexander, Sen. Patty Murray and distinguished members of the Senate Committee on Health, Education, Labor and Pensions, it is my honor to testify before you today on the reauthorization of the Elementary and Secondary Education Act (ESEA), and to share with you the perspective of a classroom teacher on how the ESEA should address the issue of testing and assessment.
I am a proud New York City public high school teacher. Currently, I teach both English and U.S. history to 11th-grade students at Harvest Collegiate High School in Manhattan, a school I helped found with a group of teachers three years ago. I also serve as our dean of Academic Progress, overseeing our school’s assessment system and supporting student learning schoolwide. My students, who are listening to us now—and who I need to remind to study for their test tomorrow—represent the full diversity of New York City. Over 70 percent receive free or reduced-price lunch; 75 percent are black and/or Latino; 25 percent have special education needs; and the overwhelming majority are immigrants or the children of immigrants.
After receiving my undergraduate degree and teacher certification at Brown University, I began my career not far from here in Fairfax County, Va., at Hayfield High School. I moved to New York 10 years ago at the behest of my then fiancée, where I completed a master’s degree in African-American studies at Columbia University. I then taught for five years at the Bronx Lab School and a year at the Academy for Young Writers in Brooklyn before starting Harvest Collegiate in 2012.
I am a National Board Certified teacher who was twice elected by my colleagues to serve as their union chapter leader. I have helped lead the development of local performance assessments in New York City to be used for teacher evaluation, developed prototype tasks for the Smarter Balanced Assessment Consortium, and am currently writing model 11-grade United States history curriculum units for the New York State Social Studies Toolkit. I speak to you today not only as a teacher who cares deeply about his students, but as one who is evidence that teachers are capable of and willing to impact the larger education world beyond our classrooms.
Most important, however, I am a teacher who got an email from Brandy in the middle of her freshman African-American Studies seminar at Lehman College complaining that without me, “there's no one around anymore to ask me the all-annoying questions about my beliefs or passions." I’m the teacher to whom Genesis wrote, “Thank you for an amazing unforgettable four years. You were the first teacher to ever be true with me and point out my flaws. You made me realize that everything we do, whether it be academics or decisions in life, has a greater purpose than we intend them to have." To whom Robert wrote, “I remember one time you told me that I had all the right tools to become a strong leader. I just want you to know if that’s true, it is because you helped install those tools." To whom Rosio wrote, “Many teachers at this school have cared for me, but Steve has been the one to take time and to look after me and guide me in the right direction since day one.” And to whom Tyree, after finally passing a state standardized test on his fourth try wrote, “Thank you so much. You pushed me to the top.” I am now proud to call Tyree, a graduate of Sarah Lawrence College, a colleague at Harvest Collegiate.
Despite these accomplishments, I am a teacher who every May, until last year, would get up and apologize to my students. I would tell them, “I have done my best job to be an excellent teacher for you up till now, but for the last month of school, I am going to turn into a bad teacher to properly prepare you for state Regents exams.” I told my students there would be no more research, no more discussion, no more dealing with complexity, no more developing as writers with voice and style. Instead, they would repeatedly write stock, formulaic essays and practice mindless repetition of facts so that they could be successful on the state Regents exams in English and history. Every year, I sacrificed at least a month of my students’ learning, and I’m sad to say, it worked. My students always performed 10-20 percent above city averages on the exams.
I did this because standardized tests measure the wrong things. I did this because the stakes for my students, and more recently for me, forced me to value three hours of testing over a year of learning and development. I did this because the standardized test was the only way for my students to demonstrate their learning to the state and federal governments. My incentives were all wrong.
Ted Sizer, the Brown and Harvard professor whose book Horace’s Compromise made me decide to become a teacher 15 years ago, emphasizes the need to get the incentives right for schools to work well. Students need clear and relevant goals. Teachers need autonomy and accountability for helping students progress toward those goals. Sizer made clear that when we get the incentives right, schools flourish.
Right now though, the federal incentives in education are wrong. Too many schools are designed, in large part, to get students to do well on a one-time test. We need to reverse that hierarchy so that schools can organize themselves primarily to help students learn. I applaud this committee for the work it has done to begin to get the incentives right. As a teacher, I hope to offer some further insight into the negative effects of our current system of testing, and offer some lessons from my school and others like it about what can work better.
Our first problem is the limited tools for assessing what students know and can do. Because of the demands of testing every student every year, and the psychometric demands of high-stakes assessments, most federally mandated tests are a one-time assessment that privilege multiple-choice questions over authentic demonstrations of students' knowledge and skills. Teachers learn little from these exams that can lead to better instruction and increased learning, especially when they come at the end of the year. As I would prepare my students for the New York State Regents exams in both English and history, I learned that a student's score on the test could shift by 15 points in either direction, depending on the version of the test I used and how the student was doing on that day. For an average student, that's a range that includes both the mastery level and failure. How a student does on one test only really tells me how the student did on that test on that day; to know anything of value, with validity and reliability, I need multiple measures over time which can help me understand what my students know and can do.
Standardized tests can only measure certain things. They work well for basic skills, such as reading comprehension and simple computations. However, to assess what students need to know and be able to do to be successful in college, career and citizenship, we need more-sophisticated assessments. Although not perfect, the Common Core, Next Generation Science standards, and the C3 Social Studies Framework clearly articulate the skills students need. To cite one example of many, Common Core Writing Standard 7 demands that students “conduct short as well as more sustained research projects to answer a question (including a self-generated question) or solve a problem.” This standard cannot be assessed through an on-demand test; rather, students will actually need to perform research over extended lengths of time and be assessed for their ability to do so.
But even if we can improve the tools, when the stakes of testing are high, students do not get what they need. In all too many cases, the test becomes the curriculum. A former colleague of mine who coaches science teachers for a well-known charter school observed teachers using the year-end test as the only motivation for students to learn science. This is science! If the only reason students believe they need to know science—or any subject, for that matter—is the test, America is in trouble.
Making the test the curriculum harms all students, but it does the most harm to those with the lowest skills. When I taught seniors in the Bronx, I worked with the highest-performing students who had already passed all their exams to help prepare them for the rigorous reading and writing they would face in college. We read philosophical and theoretical works ranging from Kant to Rawls to Nozick, and wrote and revised college-level argumentative essays. Though the Common Core was a far-off whisper at that point, my course far exceeded its demands, even if all my students could not yet meet them.
At the same time, I worked with the lowest-performing students who had yet to pass the New York State Regents. With them, I focused on mindless repetition of the facts that make up most of the Regents, and combined it with writing formulaic, timed essays that bore little resemblance to any real academic, civic, or career-based writing. I was really good at it, getting 100 percent to pass their exams in my final year leading Regents prep.
But I was doing my students no favors. I think to this day about T., a second-semester senior who could hardly write and struggled to read. In "Regents Prep Class" I worked with her on rote memorization rather than improving the reading, writing and thinking skills she would need for the rest of her life. The incentives were all wrong; sure, she passed the test, but she was still not ready for the community college work she encountered that fall. When we focus our efforts only on helping struggling students jump over the hurdle of mandated exams, the learning and opportunity gap widens.
I also think about how arbitrary it is to get a passing grade on these exams. J., a student with special needs, didn't graduate on time because he ran out of time on the test. He was one point short on a Regents exam, and still had an entire essay to go. Had he had time to write just one sentence of that essay, he would have passed and graduated. J. was ready to graduate, but because there was not one more minute for him in June, he had to wait until August to finally succeed.
Annualized tests taken at the end of the school year do not provide teachers and schools with the information we need to best serve our students. We need information about what students can do on real, authentic tasks. Can they make sense of a newspaper story, use it to inform their views, and detect the bias within so that they can become informed voters? Can they write a speech, so that, if one day they are given the chance to testify before a city council or the U.S. Senate HELP Committee, they are prepared to do so? Can they develop complex questions about the world and use historical, logical and scientific modes of analysis and research to answer them? We need actionable information about what our students know and can do. And that information is useless to me in June: I need it early in the fall, when I am plotting out the instruction my students will receive over the course of the school year.
Many have voiced concerns that to remove the annualized testing of every student means that many students, particularly those who are poor, black or brown, will fall through the cracks. These concerns must be addressed. Schools need to know how we're doing, and the parents of our students and the community at large need to know that we are successfully educating our students. In my school, we need to constantly check to make sure all our students are learning. We need to do this through analysis of individual learners and groups to ensure we are providing every one of our students with the quality education that is necessary if they are to escape poverty, establish a productive and meaningful career, and function as active and informed citizens in our democracy. But to do this well, systems need to be built at the school and district level.
Educators on the frontlines in America’s public schools can’t do this work all on our own. To educate students with the greatest needs well, we need resources and supports that only government can supply. Fifty years ago, the federal government recognized its responsibility in this regard and enacted the ESEA into law, providing crucial Title I funds for the education of students living in poverty. As Congress reauthorizes the ESEA in 2015, I ask you to keep foremost in your mind the historic equity mission of this legislation, and ensure that our schools receive the resources and supports we need to educate all of our students well.
At my current school, Harvest Collegiate, we serve students who are representative of the New York City public school population. This means our students are relatively evenly spread from being some of the best in the country to coming to our high school at second- or third-grade levels. We not only need to ensure that all students learn, but also that those students who come to us with lagging skills and knowledge are accelerated toward our high standards, even when they begin far from those standards. We do this through quarterly performance assessments, and we invest heavily in the time necessary to analyze the data, and for teachers to use it to inform future planning, intervention and differentiation. This system has helped us ensure that, on average, students make more than a year of growth in the core disciplinary skills and habits of mind we assess. It has also kept us honest as we have real, timely information about which individual students, and which groups of students, are learning less than those around them. With this information, we can reprioritize resources, professional development, and lesson plans to better meet the needs of all students, particularly those with special education needs. This system works well because Harvest Collegiate is a collaborative school with a strong ethos of trust and community, and is part of New York City’s PROSE schools, with freedom from some departmental and contractual regulations in order to adopt educational innovations.
Our formative assessment system is in large part possible because of our membership in the New York Performance Standards Consortium, a group of schools that can serve as a model for the rest of the country and must be protected in a newly authorized ESEA. The consortium consists of 48 schools that use a more rigorous assessment system than the Regents exams. Within the consortium, high-stakes assessments are not a one-shot, on-demand test, but a graduation-level performance-based assessment test, or PBAT. To graduate, consortium students complete PBATs in all four core disciplines, in addition to still taking the New York State English Regents. This is the work that is real and authentic to a discipline, and mirrors college work, giving consortium students a significant advantage over others once they enter college. Students write an analytical essay on a piece of literature for English, an argumentative social studies research paper, conduct or extend a science experiment, and demonstrate problem-solving at higher levels of mathematics. And in all areas, students are also required to defend their work orally as well as through written products. These PBATs are evaluated by at least two teachers at the students’ school as well as an outside evaluator to ensure the reliability and validity of the process.
As opposed to top-down standardized assessments that threaten teacher morale and professionalism, the consortium uses a bottom-up system, which develops teachers’ professionalism and commitment to students. Teachers develop specific tasks that arise from the overlap between the curriculum and student interests. We are supported in this work through extensive professional development at our school, which is primarily teacher run. It should be no surprise then that the teacher turnover rate for teachers with less than five years experience at consortium schools is only 15 percent, compared with 26 percent at charter high schools and 58 percent across all New York City high schools. These results are in spite of the fact that consortium teachers do more work and bear more responsibility than conventional teachers. We design more challenging curriculum and tasks for students. Right now, my students are sharing oral histories that they conducted of immigrants to the United States and comparing those with the experiences of immigrants throughout U.S. history, which they have researched. We collaborate with our peers, both in our school and across the consortium, far more than most other teachers. We also serve as external evaluators to ensure other schools maintain our high standards for students, and evaluate other schools’ assignments and work at our annual moderation study. This is the combination of autonomy and accountability that Sizer called for, which allows us to recruit and retain teachers of extraordinary quality for our students.
This leads to better results for students. The consortium graduation rate exceeds that of the overall New York City public schools. And a study conducted by Dr. Martha Foote (“Keeping Accountability Systems Accountable," Phi Beta Kappan, January 2007) shows that the consortium has “a proven record of producing graduates who go on to successful undergraduate careers." Eighty-five percent of consortium graduates attended colleges rated competitive or better according to Barron’s Profiles of American Colleges and persisted in college at rates higher than the national average. All this was accomplished despite the fact that the consortium schools’ pool of students include more students living at the poverty level, a higher percentage of Latinos and English language learners, and a higher percentage of students with lower English and math skills than the overall NYC public high school population.
While consortium graduation rates exceed NYC averages across the board, the difference is most staggering for the most-challenged populations: In consortium schools, the graduation rate for English language learners is 69.5 percent, compared with 39.7 percent citywide; the rate for students with special education needs is 50 percent, more than double the 24.7 percent citywide rate. Moreover, graduates of consortium schools are better prepared for college than their peers. For the cohort of 2008, the consortium’s persistence rate at four-year colleges was 93.3 percent, compared with 74.7 percent nationally. At two-year colleges, consortium students persisted at a rate of 83.9 percent, compared with 53.5 percent nationally. These results arise from only one structural difference between our schools and others in the city: We do real, authentic performance assessments in place of standardized tests. Models like the consortium need to be able to exist, and thrive, within any reauthorized ESEA bill.
Despite its many well-known flaws, No Child Left Behind did include some important features that should not be abandoned. Its disaggregation of student achievement data has put a much-needed spotlight on how the education of American youth is negatively affected by economic and social inequality. Growing economic disparity has now left the majority of our public school students living in or near poverty, and we clearly do not do enough to help these students overcome the challenges that their economic condition places in the way of successfully completing their education. Racial and class segregation not only continue to plague American schools, but are actually on the increase. On the whole, students attending schools with segregated poverty and high concentrations of students of color do not receive the same quality of education as other American youth. At a time when education has become an increasingly important factor in a young person’s opportunities for a better future and entry into the middle class, these stark inequalities doom far too many of our students with the greatest need to lives of economic, social and civic marginality. We cannot afford to turn a “blind eye” to that injustice.
That is why I believe that a stance that is opposed to any ESEA requirement for student assessment is misguided. We should be more careful and precise, more intelligent in our approach. We need to track how well our schools are serving our students with the greatest needs, so that states and local school districts can provide the supports and interventions struggling schools need to improve and help those students. But the current NCLB regimen of annual high-stakes standardized exams provides only crude and inadequate measures of student achievement. Basing high-stakes decisions about the futures of students, teachers and schools on such limited assessments has done great damage. I support the position of my union, the American Federation of Teachers, that in reauthorizing ESEA, Congress should remove the high stakes from mandated tests, limit the number of tests used for accountability purposes, allow schools to use more sophisticated and useful assessment tools such as performance assessments, and schedule mandated assessments at a time that they would provide useful and actionable information on the academic needs of students.
To do this requires a better balance of the federal government’s role in education with that of local decision-making. The ESEA was first enacted into law 50 years ago in an effort to address the many unmet educational needs of students living in poverty. The federal government’s role is to ensure that American students receive a high-quality education that meets their needs. It seems clear that when setting standards and evaluating success, the federal government needs to hold states, school districts and schools accountable for not perpetuating a "soft bigotry of low expectations." But federal and state governments need to recognize that the best educational decisions for students are made by those who are closest to them, those who possess the fullest and deepest understanding of their needs. Educators’ voices need to be the loudest in making the decisions of what is tested, how students are tested and when students are tested.
To assess well, we also need the support of the federal government in developing and implementing new, better assessments. While the Smarter Balanced and PARCC consortia have done excellent work to this point in developing more meaningful assessments aligned with the Common Core, the cost of these exams makes the already challenging political climate even more treacherous. We need these exams to be less expensive, and the funding to make similar assessments aligned to the Next Generation Science standards and the C3 Framework. All of these assessments should then be available as options for school communities to choose, rather than being forced upon them by federal or state mandates.
Despite their promise, the current implementation of Common Core-aligned tests has been extraordinarily uneven, with devastating consequences for students in places such as my home state of New York. Most schools, teachers and students have not been provided the supports and resources they need to reach the higher Common Core standards, but poorly designed and executed tests are still forced upon them with disastrous results for students. In California and other states that have provided the requisite supports and resources and used high-quality assessments, implementation has been more successful. The federal government should not mandate the high-stakes testing of every student in every grade, and neither should the states. School communities need flexibility and choice in the modes of assessment they choose for their students. Models such as the New York Performance Standards Consortium need to be encouraged to grow and flourish.
If we can develop a battery of better assessment tools, then the next shift I would ask this committee to consider is a rethinking of what is measured by assessments. Currently, students are scored in relation to an age-based standard. We need to shift our thinking toward a broader continuum of growth within a grade band. At my school, we get students who as ninth-graders are stronger than I was in 12th grade, and others who are reading at a third-grade level. Therefore, we assess students on a continuum that can capture their growth throughout four years of high school. We should not be satisfied when strong students meet age-based goals at the beginning of the year, nor should we expect students who begin the year years away from those same goals to meet them that year. Again, what is useful about assessments for teachers and students is the knowledge they give us about what to do next. It makes no sense to give a 15-year-old who reads at a third-grade level a 10th-grade exam; we know the student will fail. Not only is this useless, but as a recent report from Columbia University’s Teachers College points out, “for struggling students, repeatedly confronting demands for performance they cannot reach can undermine the motivation and confidence they must have to persist in school.” This is the effect of regular grade-based testing.
Our current system, in which struggling students who are not meeting standards in third grade are overwhelmingly not meeting standards in ninth grade, does not work. It makes no sense that high schools are expected to change the course of a student’s previous nine years of education in four years. NCLB actually penalizes high schools that work with struggling students for as long as it takes for them to meet standards and graduate: For school accountability, any student who takes more than four years to graduate appears on the high school’s roll as a dropout. If our goal is truly to ensure that academic achievement gaps are closed, then we need to offer students and schools the time to do so. With that time, students can actually develop the skills of problem-solving and persistence that are crucial for future success. If we shift measurement, and therefore accountability, toward growth on authentic tasks, then we can actually have a real conversation about how to make that happen for all students.
While up to this point I have focused on flexibility and a shift toward assessing student placement on a continuum, at some point we do need standardized information about how schools and districts are doing. My brothers and sisters at Ed Trust and in the civil rights community are right to be concerned that students who are not tested are not counted. We must ensure that every student is counted. But we can do this without testing every kid, every year. Under the current law, we use grade-span testing for high school. Believe me, even though the English test in my school isn't coming until junior year, every ninth- and 10th-grade teacher has it on his or her mind as well. If grade-span testing works for high school accountability—and I have yet to see a single proposal that says we need to adopt annualized high-stakes testing in high school—why isn’t it good enough for elementary and middle schools?
We could even go a step further to remove the burdensome time standardized testing takes from student learning. The National Assessment of Educational Progress (NAEP), universally considered to be the “gold standard” of educational assessment in the United States, does not test every student. Instead NAEP uses sampling, testing only a representative cross section of students, to see how well a state or a school district is performing. While getting school level data would require more students than are currently tested using NAEP, it would still be possible to use a similar representational sampling method to get key information about districts and schools. To test every student, every year, simply for the sake of school accountability is the very definition of government waste.
Sen. Alexander, Sen. Murray and the other distinguished members of this committee, my students, my colleagues, and I are all encouraged and inspired that Congress is putting serious thought into how to improve the education of all of our nation’s students. Far too much of the current political discourse around education misses the most important part of schooling: teaching and learning. When you make decisions about the role of testing and assessment with the reauthorization of ESEA, I ask you to keep your “eyes on the prize” of how your choices will affect what takes place in our nation’s classrooms. The fundamental purpose of testing and assessment is to inform and improve teaching and learning, so that every student can be successful in school. When we use testing as a high-stakes vehicle for sanctions and punishments, we undermine that purpose and harm American education. When standardized exams replace a rich curriculum in driving instruction, the quality of our education suffers. It is time to fix a broken system of testing and accountability. And it is time to do so with the inclusion of teachers’ voices in the process. Thank you for allowing me to add my voice today, and I hope a reauthorized ESEA will formalize the inclusion of teachers’ voices across the nation.
- Stephen Lazar