- Students, teachers, and parents need a clear idea of what students should learn each year, and those goals should be reasonable, rigorous, and the same for all children.
- Adults need to know whether students are reaching those targets, and to do so, we need a consistent way of measuring progress.
- Because it matters that students reach the goals, the system should connect incentives and supports with results.
We suspect that most adults in the United States agree with all these statements. But when we refer to these statements using the shorthand of standards, assessment, and accountability, we risk losing sight of the fundamental arguments that started us on this course of school improvement.
Every state and thousands of school districts have embraced the standards agenda for the same reason that Horace Mann, who championed the "common school" in the 19th century, used assessments to compare the quality of schools: This sort of comparison shines the spotlight on inequity and helps raise the achievement of all students. The public supports this agenda, with good reason. Parents and teachers know intuitively that the more we expect from children, the more they will achieve. And most agree that our expectations have been too low for too long.
What Parents, Educators, and the Public Want
We don't have to guess how teachers, parents, or the general public feel about standards. Three recent opinion surveys paint an unmistakable picture of support for this direction of school improvement.
In an August 2000 national poll of both parents and nonparents conducted for the Business Roundtable—an association of chief executive officers of leading U.S. corporations—8 of every 10 people said that raising academic standards is a move in the right direction (Business Roundtable, 2001). Three-quarters of parents and nonparents agree that students should have to pass reading and math tests to be promoted from 4th grade, even if they have passing grades in all their classes. Two-thirds agree that students should have to pass a test to earn a high school diploma, even if they have passed all their courses. Roughly 8 of 10 support such tests if students who initially fail the test receive extra help and more than one opportunity to pass. These sentiments are consistent even though those taking the poll also acknowledged that some students may have trouble passing tests and that the tests may not measure every important skill that students are learning.
Last October, Public Agenda—a nonprofit, nonpartisan policy research organization—surveyed public school parents and found that only 2 percent favor abandoning standards (Public Agenda, 2000). A majority of respondents want their schools to continue implementing standards rather than go back to the way things were before those reforms began. The support was as strong in five large cities in the midst of standards-based reform efforts—Boston, Chicago, Cleveland, Los Angeles, and New York—as it was nationally.
More than 7 of 10 parents believe that tests are useful in several ways: to help identify students who need extra attention, to cause students to pay more attention to what is being taught, and to hold schools accountable for raising achievement. Only one parent in 10 believes that teachers put too much academic pressure on students, that schools require too many standardized tests, or that the test questions are too difficult.
In January 2001, Education Week released the results of a national survey that had probed U.S. teachers' views of standards, testing, and accountability. Eighty-seven percent of surveyed teachers agree that raising standards is "very much" or "somewhat" a "move in the right direction," and 74 percent say the level of standards in their states is "about right." This survey offers strong evidence that standards-based reforms are working, with a majority of teachers reporting more students reading, writing, and meeting more challenging expectations in the classroom.
These surveys are valuable because they go beyond simple yes-or-no questions about whether raising standards is a good idea. Many teachers and parents have legitimate concerns about the focus on testing in schools. We believe that these concerns have more to do with specific issues in specific states than with a wide disagreement about using higher standards to drive school improvement. Even when pressed, teachers and the public still believe that standards, assessment, and accountability will lead to the kind of schools that we want and need.
Still, anyone who has studied education reform in this country during the past century knows that good ideas and strong public support do not guarantee successful and widespread improvement. The key is in the execution—the translation of the broad aims upon which we agree into policies and practices that play out in classrooms.
Getting Standards and Tests Right
When they are well devised and implemented, academic standards and tests, and the accountability provisions tied to them, can change the nature of teaching and learning. They can lead to a richer, more challenging curriculum. They can foster conversation and collaboration among teachers within and across schools. They can create a more productive dialogue among teachers and parents. And they can help focus everyone's attention on raising student achievement. When they are poorly devised and implemented, standards, assessments, and accountability can become a distraction and a source of frustration in schools.
The independent, bipartisan, nonprofit organization Achieve was created by U.S. governors and business leaders to help states determine the difference between high-quality and poor-quality standards, assessment, and accountability policies. During the past three years, we have worked with nearly half the states to examine their policies, compare them with the best examples from other states and nations, and offer suggestions for improvement.
We have drawn a number of conclusions from this work, which can best be expressed in terms of ongoing challenges that states face if standards are to bring about lasting improvements in U.S. classrooms.
Teachable Standards
For standards to have an impact on what goes on in the classroom, they must be teachable. Two qualities stand out in this regard: clarity and parsimony.
Clarity requires that the standards contain enough detail and precision to allow teachers, parents, and students to know what the students need to learn. Standards that are vague or unclear can be misunderstood or ignored altogether. Teachers may feel forced to turn to something else besides the standards to guide their instruction, and often they turn to the tests—thus fueling concerns that schools are teaching to the test rather than teaching a rigorous curriculum. If the standards are ambiguous, they also offer no assurance that every student is learning challenging material.
In our work with states, Achieve has found the clarity glass half full. Nearly all the standards that we have reviewed are clear and jargon-free. They avoid including expectations that cannot be measured, such as students' "enjoyment of reading," a pitfall of many earlier versions of standards.
But too often, the standards that we have reviewed tend to be imprecise and all-encompassing. Stating that students should "read literally, inferentially, and critically" gives little guidance to teachers trying to build lessons. Those standards could refer to Shakespeare's sonnets or to Julia Child's cookbooks. Such imprecision is a recipe for uneven teaching and unequal learning. Oregon, for example, is drafting new English standards in response to Achieve's benchmarking work. One of the new 4th grade standards requires students to use knowledge of the situation and setting and of a character's traits and motivation to determine the causes for that character's action. This standard focuses attention on the analysis of literary elements in a way that enriches instruction. The same is not true of standards that require students to identify the setting or to understand character motivation.
Five years ago, standards may have been left vague in deference to local control of the curriculum: Let the state set the broad goals, but leave the curriculum to local schools and educators. But that has not turned out to be very practical. As schools and students have begun to be held accountable for results, educators have demanded more specific guidance and teaching tools, and so far demand has outstripped supply.
In response, some states are getting creative. Indiana clarified its statewide standards in the 1999–2000 school year to provide more detail and guidance and is building aligned curriculum frameworks, complete with embedded assessments that districts can choose to implement. Vermont uses technology to help teachers align classroom teaching and assessment with the state standards. The state's system includes instructional planning and assessment management tools, an online teacher forum, and a database of its standards. In New York City, the United Federation of Teachers has launched a $2 million effort to help align standards and provide teachers time off to develop a bank of student work and lesson plans tied to standards.
Of course, states can go too far in providing clarity and specificity. Making standards more specific only helps if they become more focused as a result. Educators must make tough choices about the most important knowledge for all students to learn; a laundry list helps no one. Although politically expedient, the approach some states have taken—to include everything students could learn in a subject to avoid offending anyone—simply does not work. The end result of a glut of standards is the same as when the standards are ambiguous—they fail to define what all students should learn. Moreover, too many standards undermine the power of standards as common expectations and leave teachers feeling overwhelmed by the sheer volume of what needs to be taught. Many teachers will be more apt simply to cover the material rather than give it the in-depth treatment that we know benefits students.
Rigorous, Reasonable Standards
The public will continue to support standards so long as the students who reach them are prepared to succeed in the next grade, in college, and in meaningful careers. These are the tangible outcomes that parents want for their children. The kinds of literacy, math, science, and problem-solving skills that we once expected of only some students are needed by all to succeed in a fast-moving, technology-driven world.
So how high is high enough? The answer is murky, in part because little conclusive research pinpoints what all students need to learn to be prepared for life after high school. We do know, however, that most high school students aspire to attend college and that most parents expect a high school diploma to signify readiness to do college-level work. Unfortunately, the Education Trust estimates that although as many as 80 percent of today's 8th graders will attend some form of college immediately after they finish high school, on average one-third of those students (and as many as 50 percent of students from high-poverty areas) will need remediation in basic reading, writing, and math before they are ready to undertake college-level work (Haycock & Huang, 2001).
It is reasonable to expect that students who have met state standards are prepared to do college-level work without remediation, so we must ensure that standards are rigorous at every grade level. In Achieve's analysis of state standards, we have found that although some states have improved the rigor of their standards, others still have not raised standards high enough to prepare all students for college-level work or for high-skilled jobs (Education Week, 2001).
Another way to determine where to set the bar is to look at what other countries expect of their students. When compared with the standards of high-performing nations, many state standards don't match the rigor or depth of content that other nations routinely expect their students to master. In Japan, for example, students are asked to know place values to 10,000 by 3rd grade and place values up to 100 million, 1 billion, and 1 trillion by 4th grade. In some U.S. states, these expectations—met by Japanese students in two years—are spread over six years of math.
We recognize that many schools are already working hard to prepare students for the rigors of postsecondary life and that many students are struggling to meet those high expectations. That's why the pace at which states phase in these expectations needs to be reasonable and responsible. Texas and New York, for example, both set the initial passing standard on their tests low and are slowly raising it as schools and students adjust to higher standards.
Tests That Measure the Standards
- If it's not in the standards, it shouldn't be on the test. Achieve has found that most states that have developed their own tests have done a good job of ensuring that nearly everything found in the tests can be found in the standards. The tests should offer no surprises for teachers and students and should provide a clear path for teaching and learning. The same is not true when states use off-the-shelf standardized tests. Although generic tests will probably measure a small portion of a state's standards, these tests are general enough to use in many states with varying standards. They cannot measure the breadth and depth of each state's standards. Even more troubling, such tests may measure content that is not specified in some states' standards, sending mixed messages about what students should be learning. Tests based on standards that are clear, rigorous, and public are the fairest, most effective way to measure school and student performance. Tests not based on the standards are neither fair nor particularly helpful to schools or parents.
- When the standards are rich and rigorous, the tests must be as well. The most common problem Achieve has uncovered in our analysis of assessments is that even though the standards may include high-level concepts and skills, the tests often don't (Education Week, 2001). Most states use some combination of multiple-choice, open-ended, and essay questions, and we have found weak and outstanding examples of every kind of question. For example, some open-ended questions on math assessments that purport to measure problem solving require only that students plug numbers into a formula, rather than asking them to determine how to solve the problem and show how they arrived at that solution. We have also found that test makers underutilize multiple-choice questions—an important and straightforward measure of student knowledge. Tests don't need to measure only the most rudimentary skills and knowledge, but they often do.Regardless of the choice of question format, standards-based tests must tap both the breadth and the depth of content and skills in the standards. Otherwise, we risk narrowing the curriculum to only those concepts that are tested. In one state, for example, the mathematics assessments skirt nearly all of the more challenging concepts from the middle and high school standards, including slope; similarity; the Pythagorean Theorem; circles; surface area; volume; coordinate geometry; spatial relationships; symmetry; congruence; transformations; vectors; and measurement formulas for volumes, surface areas, or areas of polygons other than rectangles and triangles. What's left? Basic arithmetic, statistics, and other concepts that altogether account for only a portion of what we found in the standards are the topics of the tests.In another state, the English standards typically ask students to "identify and analyze." But the tests almost exclusively focus on the easier skill—identify. Again, such high-level skills as analysis and interpretation can and must be measured by assessments if the assessments are to provide the type of information that teachers and parents need. Although the test format is a factor, multiple-choice questions can measure more than simple concepts. We have seen numerous examples of rigorous and meaningful multiple-choice questions that probe both students' understanding of characters, story events, and themes in literature and their ability to follow directions, extract meaning, and synthesize information from nonfiction materials. For example, an item on the 10th grade English exam in Massachusetts asked, "What do 'shadows' represent throughout this essay?" The correct answer—inherited instincts—required interpretation of a symbol rather than simple recall. On Michigan's English tests, students are sometimes asked to analyze a passage on the basis of what they have read in another passage: "The character in The 'Open Window' who best fits the description of a storyteller by the author of 'The Importance of Fiction' is . . ." Tests that pay short shrift to such critical concepts and skills may inadvertently encourage schools to narrow the curriculum.
- Tests should become more challenging in each succeeding grade. If the standards grow more challenging as students move through school, it only makes sense that the tests should, too. The solid foundation of basic skills and knowledge for younger students should evolve into more complex skills for older students. Perhaps the most intriguing finding in our work has been that elementary and middle school tests, for the most part, are appropriately rigorous, yet the high school tests are often well below the level of challenge that we would expect of all students. In a few cases, the high school tests are less rigorous than the 8th grade tests. Sometimes this situation occurs because the tests omit the more challenging concepts, and sometimes the test questions are simply too easy for high school students.Just what content should be measured and how rigorous the high school tests should be are difficult issues that most states are grappling with, particularly when students will be asked to pass the tests as a graduation requirement. It's easy to see why states may de-emphasize the more challenging concepts and skills on tests that all students are asked to pass. Yet, far too many high school graduates are bound for college but academically under-prepared. We are doing them no favors by lowering the standards.
Tools Teachers Need
Common sense dictates that teachers will be asked to do more if students are asked to do more. That's undoubtedly the case with standards, assessment, and accountability. One of the biggest challenges in advancing standards-based reform is providing teachers with the training, tools, and supports they need to help all students reach high standards.
Education Week's survey (2001) reveals that teachers believe that they are not getting enough help. Fewer than half of teachers in the United States say that they have plenty of access to curriculum guides, textbooks, or other teaching materials or to specific training connected to state standards. Barely a third say that they have plenty of access to lessons or teaching units that match standards, and fewer than a fifth say that they have had plenty of training on using state test results to diagnose learning gaps. In fact, only four states let teachers know how each student performed on every multiple-choice item, and only nine states send teachers their own students' scored work on essay questions.
The report's analysis of professional development opportunities in the United States reveals that teacher training is not closely linked to standards. In our benchmarking work, we map each test question back to the standard that it is intended to measure as a starting point for evaluating that question's effectiveness. We would argue that the same should be possible for the college preparation that prospective teachers undergo and the professional development that practicing teachers experience. If we are to hold educators accountable for their work in helping students reach high standards, we have to know how the training those educators have received connects to the standards. Education Week reported, however, that whereas 38 states require teachers to participate in professional development to renew their teaching licenses, only seven of them require that some of the training be in the teachers' subject areas. Training in content areas is the least we should expect.
Some states are making professional development a priority. California has created professional development institutes for teachers of reading, English, and mathematics, with an emphasis on the subject matter covered in the state standards. The institutes provide intensive instruction to both beginning and veteran teachers, with priority going to teachers in low-performing schools.
A Fighting Chance for Students
If the standards represent the skills and knowledge that we agree are essential, we need to give students whatever assistance they need to reach them, whether it be more time, extra help, or a more focused curriculum. Inaction on this front could undermine the promise of standards-based reform.
The first question that schools need to ask is whether their curriculum is aligned with the state standards. If the standards are clear and rigorous, it will be easier to make this determination; if they are too broad or vague, any curriculum could be said to match. Growing evidence suggests that exposure to rigorous material pays off. Students who take a full set of college preparatory courses perform better on the National Assessment of Educational Progress (NEAP) tests in math and science and on the SAT than students who take two years or less of math (Haycock & Huang, 2001). The University of Massachusetts at Boston found that minority students in that state are underrepresented in the math classes that prepare students for the state's high school test. Such disparities explain much of the achievement gap between minority and white students on the state test (Coleman, 2001). Although a few states are responding to the cry for help, not enough is being done to identify or produce curriculum materials tied to standards.
In addition to a strong curriculum, all schools need an intervention and support system for students who fall behind. States and districts are beginning to put such programs in place, but the pace in many places is too slow. Maryland, for example, has recognized the importance of this challenge and has made addressing it a top priority. A new initiative will require schools to provide additional help to students who fall behind in reading and math, either in school, after school, on weekends, or over the summer. The state will supply the financial resources to carry out this mandate. The state will also help districts train their teachers to diagnose learning problems and provide the right help.
As for the tests, they shouldn't come at a moment's notice or be a single make-or-break experience. Students and schools should have adequate time to prepare before the tests count. Many states have followed the approach of allowing several years before stakes are attached to test results. Students should also have multiple opportunities to retake tests if they don't succeed on the first try, preferably after receiving additional targeted help.
Staying the Course
None of our observations about needed improvements in states' standards, assessment, or accountability policies should be confused with the chorus from critics who question the need for testing and accountability altogether. Some would have us abandon testing or put accountability on hold indefinitely until the tests are perfect. That would be a mistake. Standards and tests are better today than they have ever been. They can get better, but we shouldn't let the ideal be the enemy of what's very good and effective. Students have told us that they can do more if schools expect it of them. Teachers have told us that standards, assessment, and accountability are leading them to ask more from their students and that students are stepping up to the new demands. Although shortcomings still need to be addressed, we are already seeing meaningful improvements.
It is no surprise that the states making the greatest gains in reading and mathematics on the NAEP in the 1990s—such as Connecticut, Kentucky, North Carolina, and Texas—also were early and consistent supporters of state standards and tests. They have stayed the course over time, making adaptations and improvements as necessary.
Educators praise the impact of standards, particularly in high-poverty schools. When Virginia began its state tests, fewer than 10 percent of students at Tidewater Park Elementary School in Norfolk passed the math, science, or history tests. Three years later, about 8 of 10 Tidewater Park students pass the tests. And there are many other examples of high-poverty schools and districts that are using standards and achieving both equity and excellence.
But we must move from the isolated pockets of excellence to excellence for all. If educators and policymakers can address the issues above, we will see schools improve in unprecedented ways. We owe it to our children to get this right.
