Yet none of our participants suggested eliminating all standardized tests. Instead, the general recommendation was to use fewer multiple choice-type tests and supplement these with alternative performance measures that respect diverse curriculums, preserve teachers' creativity and passion, and allow for student individuality.
No Child Left Behind was applauded only for its call to improve educational quality for all underserved children. Discussing possible NCLB reforms led to a shared suggestion that federal and state responsibilities might be swapped: broader national testing standards could be developed at the federal level, while leaving states free to create their own remedies and sanctions for poor test results.
Posted February 11, 2008, 10:00am
A dozen years ago, complaints about schools giving diplomas to youngsters with pitiable skills, virtually consigning them to poverty and privation, erupted spontaneously in nearly every focus group we held. Now, people are more likely to complain about testing. To me, that’s an improvement, however mixed the results. Still, we do see a growing sense that testing has become the tail that wags the dog. In a recent survey of first-year teachers, too much testing and not enough freedom to be creative beat out low salaries and unmotivated students as the top drawback to teaching. The criticisms we hear are often tangled:
- Some center merely on the amount of testing.
- Some center on whether test results are good measures of student, teacher or school performance.
- Some center on how testing shapes curricula and how teachers spend time in class.
Just to start us off, I’d be interested in your thoughts on those three issues. Is there too much testing? Are we using it to measure the right things? And is it having a positive or negative impact on what happens in the classroom?
Posted February 11, 2008, 11:45am
Overall there’s too much testing because we have external tests coming in from too many directions, too many separate tests, too many kinds of tests, too little coordination among them. This needs to be rationalized and simplified—and could also be way shortened (as a time consumer) if everyone moved to something like computer adaptive testing.
Are we testing the right things? First of all, we should be putting a premium on more things than math and reading skills, and too large a fraction of current accountability-style testing is oriented exclusively to those. NCLB (No Child Left Behind) has made this worse. We really ought to be testing all the subjects that we think matter greatly, which for my money means (at minimum) adding science and history to reading and math. Let’s acknowledge, though, that while sophisticated multiple-choice style tests, closely aligned with very good standards (which most states lack), can “measure” quite a lot of what’s important in education, there are a number of other things we value in education that simply don’t lend themselves to multiple-choice testing (or, sometimes, to any form of testing).
Positive/negative impact in the classroom? Most of the complaining comes from teachers and from middle class parents, not from low income parents whose kids most desperately need to learn reading and math skills? Part of the complaining from teachers arises from their (understandable) preference for minimizing external scrutiny and accountability—understandable, yes, but I don’t have a whole lot of sympathy for this view. Regarding the middle class parents, they are largely grumbling about the priority given to reading and math skills, for which testing serves as external “discipline”, whereas they want a broader curriculum for their kids. So do I—for all kids. Which argues for testing more subjects—but also less total testing, shorter tests and more sophisticated tests.
Despite all the grumping, I don’t see the standards-testing-accountability regime going away and I don’t want it to. We need to work harder at getting its elements right, tho.
Posted February 11, 2008, 1:30pm
The real problem is not that we use the standardized tests to measure the performance of our schools, it’s that we do not use any other metrics. Difficult as it is to measure academic skills with any sort of nuance, it is even more difficult to measure the other important aspects of a school’s performance or culture. The solution, then, is not to stop measuring academics – data is important both to determine how a school is doing for purposes of accountability and also to identify which practices are truly effective in order to learn from them – but rather, to take on the very difficult task of measuring the other aspects that are critical to determining the health of a school.
Imagine defining a person’s physical health exclusively by the pace of her pulse, or the health of a nation’s economy solely by its inflation rate. These are certainly important indicators that we must continuously monitor, but from them we cannot determine all we need to know about the health of a person or the nation. A pulse rate won’t tell you if you have cancer, and the inflation rate won’t by itself tell you whether we’re in a recession. Similarly, we do need to know how a school is doing academically, but to have a true gauge of its health we also need to know whether its students are experiencing long-term success, whether they graduate with the intention of creating a better world, and whether they feel physically and emotionally safe in their schools (among others).
As Mr. Finn said, a good start would be measuring all the academic subjects we value, but beyond that, we need to take on the difficult but essential task of defining what else we consider important.
Posted February 11, 2008, 2:20pm
Testing has gotten out of hand. There is too much testing and the results are being misused to make consequential decisions for students, teachers, and principals.
The scores have become the be-all and end-all of education. Anyone who understands the process of teaching and learning recognizes that test scores are only one aspect of education.
Districts are now obsessed with “data-driven-decision-making.” The scores are used to promote or flunk students, to award merit pay to teachers and bonuses to principals. The scores matter only in reading and math because other subjects don’t count.
The multiple-choice tests that are now in common use are pathetic measures of educational quality. We need tests that incorporate essays, demonstrations, and student explanations of how they arrived at a solution.
Testing is not a substitute for instruction or curriculum. The current testing regime contributes to the dumbing-down of American education.
Posted February 11, 2008, 2:30pm
I agree with Chester that, as currently administered, there seems to be too much testing, that takes too long, lacks sophistication, and puts a premium on too few subjects. NCLB forced school districts and states to become more business-like in their collection and use of data to manage resources and drive performance. The challenge going forward is to design and implement testing policies which achieve the efficiencies and balance Chester yearns for.
I don’t think we are testing the right things. Nor do I think testing alone can give us all the data we need to assess whether a student is learning or whether a teacher is of high quality. How do we measure and reward students for their ability to think creatively, communicate effectively, and behave respectfully? We also need to recognize that quality teaching can’t be measured by test scores alone. It’s about motivating kids to come to and to stay in school, and getting them to value and embrace the lifelong benefits of learning. I like Ryan’s image of the “health” of a school. Why have we become so obsessed with test results to the exclusion of these other indicators of student and school well-being? I agree, Ryan, that the essential and difficult task is to define what else we consider important to a good education. But where does that defining take place-in Congress, in individual communities or at the classroom level?
Posted February 11, 2008, 3:15pm
Thank you all for your comments. They’ve been fascinating, although we are all being urged to keep our responses short, about 150 words or so.
I’ve been wondering about this: According to North Carolina’s Blue Ribbon Commission on Testing, a third grader in the state could take more than 30 standardized tests before graduating. To reduce that number, the Commission just announced 27 ideas including:
- Keeping end-of-grade tests in math and reading every year between 3rd and 8th grades
- Keeping end-of-course tests in five high school subjects
- Dropping standardized writing tests given in grades 4, 7, and 10
- Dropping an 8th grade computer literacy test
- Dropping high school tests in physical science, chemistry and physics
According to the AP story, the aim is to develop tests based on a “rigorous curriculum with subject matter that’s focused only on what’s absolutely necessary for students to be ready to enter college or the high-tech work force.” The Commission also wants tests that offer better information for teachers on a student’s academic shortfalls.
Heading in the right direction or the wrong direction? Easier said than done?
Posted February 11, 2008, 4:45pm
I say the right direction. The state is being pro-active in trying to improve its testing program. I understand the State’s Board of Education convened the Commission, which is comprised of teachers, principals, superintendents, legislators and testing experts. A refreshingly common sense approach. The Commission is recommending fewer tests and more in-depth testing that will give teachers diagnostic information which can be used to identify and address students’ academic shortfalls. I also like their effort to align testing with a rigorous relevant curriculum, focused on graduating students with the skills they will need to be ready for the college or workforce.
Easier said than done? Probably. No matter how blue the Commission’s ribbon is, there will be those who will disagree with its recommendations. But the process and direction of change seem right.
Posted February 11, 2008, 6:00pm
As a classroom teacher I think the real problem with how test scores are driving reform is this—all students are being measured against one uniform set of high standards. (I know this uniformity is easier to justify and simpler to explain to the public.) Kids are not widgets—every student comes to school as an individual with their own history, culture, strengths and weaknesses. Standardized tests appear to be trying to bring a level of quality control to schools similar to what would occur in a factory setting where the products are all identical. Every child is unique, and we need to recognize that as we work to develop tools to assess improvement and growth in every student. The goal should be to take each student from their current level of functioning and go as far as possible with them. (Is it failure to take a sixth grade student who reads at the first grade level in September and have them reading at the fourth grade level by May? In the world of standardized testing it is.)
Posted February 11, 2008, 6:45pm
Whether this is a good move or not depends on a number of factors, not least of which is the quality of the tests. Thirty tests sounds like a lot, but whether that is an excessive number depends entirely on how good the tests are, and how those tests are used. I share Diane’s (and others’) concerns with over-testing or misusing narrowly-focused tests, but I also get nervous when states de-emphasize important subjects like computer literacy and writing, as these skills are becoming increasingly important and decreasingly well-developed by schools whose focus has become too narrow. As the saying goes, “that which is measured is managed”, and I’d hate to see schools stop teaching critical skills like computer literacy and writing, which is inevitably what some schools will do if they are no longer measured according to their success in teaching them. “Absolutely necessary” indeed; what could possibly be more necessary in the information age than computer literacy?
Data-driven decision-making, like any management tool, is only as good as the manager using it. To a good manager, good data can be extraordinarily positive, allowing for a deep discussion of what is working and what isn’t, who is learning and who is not, which skills the class has mastered and which need to be re-taught. Data need not remove the humanity from teaching, as they can take the form of narrowly-constructed multiple-choice tests, well-designed tests that combine multiple-choice and writing, or even expansive and thoughtful rubrics that score a student’s performance on a multi-disciplinary project. Data need not be constricting or bluntly applied.
Posted February 11, 2008, 11:30pm
We are in the midst of an era of excessive standardized testing, with three defining characteristics.
First, as many of the participants have already noted, there is an excessive number of standardized exams. The state accountability system requires one set of exams under NCLB, and far too often the local school district adds its own exams for political reasons. Diagnostic tests and formative assessments are layered over exams that test for mastery of critical skills, of specific subject matter and even of general intellectual aptitude. Here in NYC, we now have “interim assessments” every six to eight weeks, out of a misguided notion that such perpetual testing will provide a constant stream of feedback on the progress of students. As the sheer quantity of standardized tests grows, the proportion of quality standardized tests diminish: the sad truth is that today far too many state education departments and school districts construct such exams “on the cheap,” or buy them from testing corporations that place profit before quality.
Second, there is widespread misuse of the standardized exams that do exist. Testing experts – psychologists and psychometricians – design and construct quality tests for very specific purposes. When a test designed for one purpose, such as the diagnosis of the problems individual students face in developing reading comprehension, are used for a very different purpose, such as measuring those students’ mastery of literacy skills, the results are no longer valid and reliable. Here in NYC, there is now talk of using state exams designed to test student mastery of foundational literacy and numeracy skills to evaluate the classroom performance of teachers – as stunning a misuse of standardized exams as one could imagine.
Third, the professional associations of those who design and construct tests warn that the greater the stakes attached to an exam, the less reliable its results become. Yet school districts continue to place more extraordinarily high stakes on the outcomes of standardized exams, making decisions about a student’s promotion and graduation on the basis of a single test. One of the reasons why the NAEP exams stand out in the field, with high rates of reliability and validity, is that they are low stakes tests: there are no punitive consequences for low scores. The problem of test anxiety is minimized, and there is no “test preparation” to raise scores.
Posted February 12, 2008, 10:00am
Some districts are beginning to recognize that they are overdosing students in standardized tests, but most are not. I was amused to see that the typical third grader in North Carolina “could take more than 30 standardized tests before graduating.” Ha! In New York City, the typical third grader will take 10-12 standardized tests in third grade alone!
As you note, North Carolina is re-examining its standardized testing schedule. I was particularly impressed by this comment by Sam Houston, who led the commission to re-examine North Carolina’s testing requirements: “We’re testing more but we’re not seeing the results,” said Sam Houston, the commission’s chairman. “We’re not seeing graduation rates increasing. We’re not seeing remediation rates decreasing. Somewhere along the way testing isn’t aligning with excellence.”
Florida is also re-thinking its heavy reliance on the FCAT. Here is a recent editorial in the St. Peterburg Times.
I agree with Sam Houston. The tests we have now, the subjects we test, and the way we test are not producing excellence. They may or may not be raising the floor, but they are very likely lowering the ceiling. They are certainly not conducive to educating the leaders of the future.
Day Two of Discussion
Posted February 12, 2008, 10:15am
It appears that while standardized tests appease the general public, everyone actually involved with education recognizes and agrees that there must be a better way to improve quality and accountability.
High stakes tests have a negative effect on all the parties involved. Students who believe they cannot pass a test required for promotion or graduation simply drop out. (This may initially look good on paper, since those potentially failing scores are never recorded.) High quality teachers in low performing schools jump ship to move to more affluent, high-performing schools. (This results in inexperienced teachers being in charge of the students with the highest needs.)
I may be naive, but I do not believe that the majority of teachers want to avoid accountability. We do need to develop more flexible and adaptable tools, and we need to educate the public on the real complexity of the situation.
There is an old saying- A pig can’t gain any weight if it spends all its time on the scale. All that testing means a lot less time for learning.
Posted February 12, 2008, 11:00am
I am struck by the fairly wide agreement that we have too much testing, and that current tests aren’t really functioning all that well in insuring rich, rigorous curricula. And I don’t really hear anyone suggesting that we can do a good job in public education without tests. So what would be needed to get genuinely high quality tests? Should they be national, perhaps growing out of the NAEP tests? Are annual end-of-the-year tests sufficient? These wouldn’t provide much help for teachers in terms of diagnostics. If you had your dream solution for developing high quality tests and testing plans, what would it be?
Posted February 12, 2008, 1:00pm
Tests are important in evaluating the effectiveness of education, but they are not the only tool available. (Tests are quick, generally cost effective and don’t need much explanation.) We are limiting ourselves, and our students, by focusing only on tests. Other mechanisms for evaluation will take more time and effort on the part of educators, and more explanation to parents and community members, but are so much more effective.
Other forms of accountability include Authentic Assessments or Performance Assessments: major assignments where students must produce a product. For example in physical science—design and carry out a science experiment including control of variables, gathering and presenting data and drawing conclusions based on that data. In English, an example would be responding to a novel in newspaper format. The student must write a hard news story, a gossip column, and their choice of an obituary or a political cartoon to demonstrate understanding.
A specialized rubric is created for each assessment so that grading is as standardized as possible and students know exactly what is expected of them.
These assessments take more time to design and grade and they may take more time for students to complete. In the end, however, the teacher has a much clearer picture of what the student actually knows!
Posted February 12, 2008, 1:26pm
I agree with the comment by Deb White. If schools really want to test whether students know and can do, then they must use assessments that ask students to demonstrate their knowledge, not just to pick one of four bubbles. The standardized, multiple-choice test is quick and easy and cheap to score, but it degrades learning when it is the only measure. If we expect excellence, we need better measures of learning.
If we expect minimum competence and mediocrity, then we are on the right track.
End of course exams that ask students to demonstrate their mastery of the subject are appropriate and are administered at the right time.
Posted February 12, 2008, 3:48pm
We need to distinguish between the various uses of standardized tests. For the purpose of school evaluation (as opposed to other possible uses, like student promotion/retention decisions), KIPP contracts British school inspectors to work with KIPP employees and school leaders from other regions in performing robust school inspections that focus on the quality of teaching and learning they see in the classrooms, the degree of student achievement they see, the quality of both operational and instructional leadership, and a number of other factors. From this inspection (which takes multiple days), you get a real sense of the quality of the schools, and the training and experience of the inspectors lead to reliable and consistent analyses.
Standardized test scores are definitely a part of the picture, but so are other, more qualitative judgments: the attitudes the students express toward their own schooling, the levels of innovation seen in the classrooms, the quality of the curriculum, the degree of critical thought demonstrated in student conversations, the degree of equity in the performance of different demographic groups, and so on. This type of inspection is fairly expensive and time-consuming, but it certainly accounts for many of the concerns we have with the limitations of the feedback provided by standardized tests as they are currently conceived.
Posted February 12, 2008, 5:20pm
It is interesting that KIPP uses British inspectors—maybe you’d like to say a little more about why you do. I am wondering whether one of the issues behind the over- reliance on standardized tests (versus subtler forms of evaluation) is that so many reformers are skeptical about whether teachers can and will make rigorous judgments about student learning.
Does the group believe that the current reliance on standardized tests is mainly a matter of convenience and cost? Or is there a sense that standardized tests are better because they are less dependent on the judgments of educators. Is this a legitimate concern? Are there ways to structure other forms of evaluation to address this? What could be done to enhance teachers’ credibility as rigorous judges of whether students are developing skills or not?
Posted February 12, 2008, 8:40pm
I can’t speak for the decisions KIPP makes at the national level, but having gone through a few inspections both as a principal and now as an executive director, I can vouch for the quality of the experience. Oftentimes, state-run charter school inspections in the United States are focused on compliance more than results or quality of instruction. At the federal level, NCLB focuses both on compliance (in areas like Highly Qualified Teacher certification) and results (Adequate Yearly Progress), but the only results it examines are test scores. As we’ve discussed, standardized test scores are important, but they don’t give you a complete picture of the health of the school. KIPP is able to work with the British inspectors to design an inspection that focuses on all the outcomes that KIPP considers important, in addition to analyzing those aspects of the school that indicate whether the outcomes will likely continue to be successful. The feedback we get from these inspections is of a very high quality and is actually useful to the school, and is geared more toward KIPP’s rigorous performance standards than to the minimal proficiency standards to which schools are often held.
Posted February 13, 2008, 12:00am
I am returning to the question of testing and teacher credibility. Standardized tests do give a sense of comfort to many reformers because the results are so concrete and “improvement” is so easy to measure.
It is interesting to me that the public trusts teachers with their children for hours every day, but then has little confidence in the ability of those same individuals to decide if their students are actually learning.
We have come full circle. The absolute best teachers I know are creative, innovative and flexible. Those are the very qualities crushed out of the school day by over- testing. As I mentioned in my previous post, there ARE ways to structure evaluations that both support creativity and joy while still maintaining consistency in grading and scoring.
There are also ways to enhance teachers’ credibility as rigorous judges of student learning. This would be easier to do if teachers were not in the double bind of having their job security depend on how their students perform. For example, a teacher with a large percentage of low functioning students, knowing that her job evaluation depends on how those kids score on the “big” test, would find it far too easy to “teach to the test” …obviously a bad situation.
Posted February 13, 2008, 12:00am
Standardized exams are only one assessment tool. For some assessment tasks, such as the testing of fundamental literacy skills, a properly designed and correctly used standardized test can perform well. Similarly, a quality diagnostic exam can be of invaluable assistance in identifying the specific problems a struggling student is having in learning how to read.
But for other tasks, standardized tests are a poor proxy for better assessment tools. Take one essential skill for success in post-secondary education and in the 21st century knowledge economy—the ability to craft a well-written, persuasive essay, which marshals logical arguments and appropriate evidence in support of a thesis. This skill requires a performance assessment in which the student must actually write the essay.
As a general rule, the more complex the skill or knowledge that is being assessed, the more limited the utility of standardized exams in assessing it. A quality education will thus include standardized exams, but as only one part of a full regimen of assessment. A rigorous course of study and a comprehensive curriculum will necessarily employ many performance assessments.
The use of performance assessments is an issue only because the process of grading an essay does not have the same aura of “objectivity” that marking a multiple-choice test does. But it is well worth remembering here the sage advice of one our greatest figures in the field of hard, objective science. “Everything that can be counted does not necessarily count,” Einstein wrote, and “everything that counts cannot necessarily be counted.” What better starting point for our understanding of assessment?
Posted February 13, 2008, 12:02am
I believe that districts and states rely heavily on standardized, multiple-choice exams for several reasons: First, they are the cheapest form of testing as compared to any sort of test that must be read by a person. Second, they are considered objective, in the sense that the questions and the answers are the same for everyone. And third, I think that many non-educators (and some educators) don’t trust teachers’ judgment, believe that they will inflate their students’ scores so that they look better, and believe that teachers are opposed to any kind of accountability. Teacher evaluations, in short, are not reliable, as compared to a machine-scored test.
Thus the dilemma: the machine-scored test may be objective and reliable, but the machine-scored test has a deleterious influence on what is taught and what is learned. The tests, when overused and misused, may actually be harmful to the overall goals of education. So, what is cheap, easy, and objective may also be toxic.
Posted February 13, 2008, 10:40am
There seems to be a consensus here that testing policies need to be improved, and here we are in an election year. So far, education hasn’t been a major issue, but let’s say that you’re advising one of the Presidential candidates. What kind of specific proposals would you suggest he or she adopt on No Child Left Behind and the use of standardized testing?
I think it’s fair to say that right now most specific testing decisions are in the hands of the states and districts. Do you think a more cohesive national standards and testing policy would be better? Or, if you believe the bulk of the decision-making in this area should be left to the states, what specific changes would you recommend?
Posted February 13, 2008, 10:52am
I am intrigued by how the issue of trust has surfaced in our dialogue. Deb points out the irony of parents giving teachers responsibility for their children, but not trusting them to be honest or fair when it comes to evaluation. And the irony of NCLB, by making teachers feel their job security depends solely on test results, the double-bind is crippling. Teaching to the test isn’t real teaching or real learning.
The current policy debate seems very bogged down by excessive focus on testing. Witness the recent back and forth between the UFT and DOE on use of test scores to “secretly” evaluate teacher performance. Too many worthy innovative ideas, like testing an alternative school year, get shot down before they are even piloted, because fundamentally the unions and management don’t trust one another. We need to move away from this excessive focus on testing, and face the cost and objectivity issues associated with different measures of student learning and teacher quality, and at the same time take a hard look at what needs to be done to restore trust to school communities. Leo, I’d particularly welcome your thoughts on the importance of trust in school systems.
Posted February 13, 2008, 11:29am
The right way to fix NCLB (and my advice to the next President on this front) is to develop a competent set of national standards and tests and cut scores by which all kids/schools/districts/states will have their performance monitored and reported and compared. Then FREE UP states, districts, schools and parents to improve their schools (or change schools, etc..) as they see fit. Current NCLB is far too prescriptive (yet ineffectual) regarding how to FIX low performing schools, yet far too laid back about the standards, tests and cut scores by which performance is deemed to be satisfactory (or not). NCLB 2.0 should turn this 180 degrees around, be prescriptive regarding ends and loose regarding means.
How to set those national standards and tests? Not Congress. Not the Education Department. Private funding (e.g. Gates) should make a whopping grant to a consortium of Achieve, the National Governors Association and the Council of Chief State School Officers to do this. Their product would be voluntary for states to use or not—and some states would stick by their own versions, at least for now. I see no way to make it compulsory for all states. NAEP would remain the external monitor and barometer by which the performance of ALL states (and nation) could be compared whether they use the “national” standards/tests or not.
Posted February 13, 2008, 11:36am
My proposal for the reauthorization of NCLB is this: the feds and the states should switch roles.
Under current law, the states do the standards and testing, and federal law prescribes specific sanctions that are supposed to “fix” or “reform” schools: choice, tutoring, reconstitution, closing, state takeovers, etc. None of this is working.
The states are setting standards and cut-points for tests that vary all over the lot (see the Thomas B. Fordham report called “The Proficiency Illusion”). So the public is not even getting accurate information about how schools are doing. Meanwhile, the federal “remedies” are ignored or ineffective. Many studies have shown that only a tiny percentage of students get “choice” to a better school (turns out the better schools don’t have many or any empty seats), and the tutoring business has boomed but with mixed results and uncertain beneficiaries.
Bottom line is that improvements as measured by NAEP are small at best, certainly smaller than in the few years preceding the adoption of NCLB.
Therefore I hope that we will end up in the next iteration with national standards and national testing (run either by the federal government or by a federally-funded private entity or by a consortium of states or something else), so that we have accurate information about performance and progress. And I hope that the sanctions, remedies, and strategies for change are left in the hands of states, districts, and those closest to the schools.
Posted February 13, 2008, 12:50pm
I agree with Diane’s comments regarding switching the Federal and State roles. I don’t believe that most teachers would have an issue with a national minimum set of standards, as long as there was flexibility in demonstrating their attainment. “Local Control” is always an issue with education reform. I think communities and school boards would be willing to relinquish some control if the system which was created remained flexible and adaptable rather than appearing punitive.
Our society is dependent on graduating students with a variety of skills. We need all kids to meet the minimum standards necessary to be a functioning member of society and then blossom in their areas of interest.
It appears that we all agree that standardized tests are just one piece of the assessment puzzle. A necessary evil that should play a much smaller role than it does currently. The challenge is to make the public, and politicians in particular, understand the limitations of testing and encourage alternate forms of accountability.
Where do we go from here? As NCLB is re-written or replaced, my pragmatic suggestions would be to:
- Educate all parties involved
- Allow multiple opportunities and venues for success. (Be able to demonstrate proficiency in a variety of modalities.)
- Create a menu of assessment tools which states or districts could choose from to evaluate their students. (This would maintain a large degree of local control.)
- Train teachers to create assessments and rubrics which are appropriate and interesting to their students (Wyoming has a consortium devoted to developing Body of Evidence assessments and corresponding rubrics.)
- Train teams of teachers and administrators to carry out school visits where school climate and other more subjective components can be evaluated. (i.e. KIPP model) Groups could swap evaluations with nearby schools to reduce bias and increase objectivity while controlling cost.
- Include representation from all levels in the education system, including classroom teachers, in all of the above. (Without teacher buy-in, any change will be difficult.)
The problem is that all of these suggestions will take more time and cost more money than administering and scoring a multiple choice test.
At the end of the day we need to remember that everyone involved in education is there because they want what is best for kids. Anyone in it “for the money” quickly realizes there are easier (though less rewarding) ways to make a living and moves on.
Posted February 13, 2008, 3:10pm
Trust is a key element in education. School communities that are successful are distinguished, in significant measure, by high levels of trust—between the adults and the young people in the school, between the school and the community, especially parents, and among the adults within the school community.
Trust is built on the foundation of strong relationships. When two parties are working together, in collaboration and partnership, that work builds trust. You don’t set out to build trust; you set out to work together towards a common goal, and trust results, leading to more common work, in a virtuous cycle.
The problems we have in New York City have their root in an unilateral, top-down vision of change—revolution from above, imposed on those who do the real work of educating children. NCLB suffers from a similar fault. If there were a real partnership around an issue like the value-added pilot project, it would never have taken such a flawed form. In this regard, the lack of trust is both symptom and cause, and the result is a downward spiral.
Posted February 13, 2008, 4:00pm
Leo is exactly right. Top down change is rarely successful or long lasting. Get all of the interested parties together (including classroom teachers). Set a common goal of increased student achievement and and success. Watch real change happen.
Posted February 13, 2008, 7:16pm
We’re nearing the end of our discussion, and for my part at least, I have found the exchanges genuinely interesting and thoughtful. I would like to thank all of you for your ideas and your willingness to participate.
As a closer, I am curious about your own reactions to the conversation. Did it make you more hopeful that the testing situation can be improved? Did you see any unexpected areas of agreement? Were there unexpected points of tension? Did anyone raise ideas or concerns you hadn’t thought much about before?
Personally, I hope this might serve as an opening gambit for a longer and broader conversation, and we would all appreciate any suggestions you have for ways to improve it. Are there additional perspectives that should be brought into the mix? Are there groups or venues that might be interested in building on the dialogue here and moving it forward?
Again, thank you. Hope our paths will cross again sometime soon
Posted February 13, 2008, 7:39pm
Thanks to you, Jean, for asking good questions, like a first-rate dinner host, and keeping the conversation going. As you hoped, people responded quickly and from the heart.
I was surprised to learn that there was something approaching unanimity on the question of whether there is too much testing. I also think that the discussants recognize that NCLB as presently written is not adequate to move this nation forward. Possibly other questions might have shown divisions, but the ones you asked did not.
Yet the very unanamity in our discussion reminded me that Congress seems not to be listening. From what I have heard and read, they seem oblivious to public opinion or the opinion of the educators in our discussion. They are barreling ahead, or so it appears, as though everything in NCLB were just fine.
There is a disconnect, and perhaps it will take a new President to shake Congress up, to change the discussion on Capitol Hill. I am pleased that thoughtful people are in agreement. I am despairing about whether our policymakers are listening.
Posted February 13, 2008, 7:48pm
I generally agree with Chester’s and Diane’s remedies, especially if we assume constant funding levels. The tests and standards absolutely should be established at the federal level and applied nationally. If we can increase funding for the assessment part of NCLB, we should be looking into more sophisticated means of measuring like those suggested by Deb. I also think Leo and Deb are on the money about the difficulty of effectively changing a system from the top-down – everything works better with as much buy-in as possible among all the stakeholders.
In a perfect system, entire districts would do something like what we at TEAM did last summer – work with the teachers and principals to develop a shared vision and accountability system (the outcomes I described earlier) and then collaborate in developing a strategy to implement it. Political realities make this sort of bottom-up approach more difficult in a district than in a small network of charter schools, though: it takes a lot of time, which doesn’t fit into the timetable of the political cycle or the urgency with which the system must be fixed; the teachers in our schools are hand-picked according to their mission-fit, so there is a built-in level of agreement going into the process; and as a school of choice we have a certain amount of immunity from the sort of political process that can make true reform efforts die a death by a thousand cuts. In short, the fact that teachers and parents choose our schools in part because they believe in its mission means that there’s enough like-mindedness to prevent the sort of battles you’d likely see if you tried building an accountability system involving all the stakeholders in, say, New York City. (In fact, when I was a teacher in NYC, I participated in just such an attempt to develop an accountability plan for our school – it failed miserably in part because of the lack of a shared sense of the school’s purpose and the amount of time it would take to implement the process correctly.) In a gigantic city where the average superintendent tenure has been just a couple of years, it’s hard to imagine anything short of a dramatic and immediate shakeup doing the job.
Back to fixing NCLB: It needs to be said that the best thing about NCLB is the extremely important message it sends that it is no longer OK for schools to fail to educate any subgroup – be they low-income, special needs-classified, or members of any other demographic group that has historically been so underserved by our education system. For districts like the ones that KIPP students come from, this is an absolutely revolutionary statement. There are problems with it, though. The rule that every single student must pass the tests by 2014 – while a laudable aspiration – is simply unrealistic (a term I as a KIPP principal am loathe to use because of its wide-spread historical usage in excusing the under-performance of schools serving kids like those we serve). Rules around Highly Qualified Teacher certification are too process-heavy in most states and do not adequately assess a teacher’s actual teaching ability – I’ve seen a number of “highly-qualified” teachers who shouldn’t be in a classroom, while some of the best teachers I know were not “highly qualified” when that rule came into effect. These two significant flaws make it easy for critics to write off the whole law as ridiculous, weakening its authority as an important piece of civil rights legislation.
Posted February 13, 2008, 10:11pm
At the end of this brief exchange, I am left with two observations.
First, as Diane Ravitch points out, it is striking how much common ground our diverse group found on the subject of standardized testing. There is broad agreement, it would seem, that our schools are being burdened with excessive testing and that as a consequence, important teaching and learning is being crowded out.
Second, it is interesting how quickly a discussion of testing branches out into other educational issues. The issue of testing, and what has gone wrong with the regimen of testing we now have, is at the center of American education. In many ways, standardized testing as we now know it defines the educational era in which we find ourselves. In significant part, therefore, the path to our educational future lies in figuring our way out of our current testing thicket.
Posted February 14, 2008, 12:05am
This entire exchange has been both interesting and encouraging. I now have some hope that a concerned and educated group could develop a workable solution if given the opportunity and the resources.
Thank you for the opportunity to represent classroom teachers. -Now I need to go and grade some papers.
Posted February 14, 2008, 3:36pm
So how do we get Congress, Presidential candidates, and ultimately the public away from the easily digestible, but overly simplistic rhetoric in support of sticking with NCLB and standardized testing: “we need tough standards, we need to increase not dilute accountability, we need to stop short-changing underserved children, look at the test scores-we keep falling behind the rest of the world and that’s why our economy is weakening.” This language only distracts us from actually tackling head on the very real policy challenges we’ve been discussing.
Two key audiences need to be better informed about the pitfalls and more engaged with educators in solutions: the business community and the civil rights community. Their support for NCLB and standardized testing is understandable. Yes, America needs to do a much better job at preparing its future workforce and at narrowing the gap between whites and students of color. But NCLB isn’t the path to achieving these goals. Is our civic culture so broken that civil rights advocates, business leaders, and educators can’t find common ground upon which to engage in more constructive policy making?
I recommend taking a look at an interesting article by Peter Henry, a Minnesota English Teacher, titled, “The Case Against Standardized Testing” (PDF). Henry makes a compelling case for the harms that flow from the high-stakes, measurable accountability movement. It’s an interesting read in light of our conversation; I found his “Dirty Dozen” quite compelling.
Thanks for an enlightening few days of conversation.