Archives for category: Testing

Fred Smith, a testing expert who worked many years at the New York City Board of Education, has become a mentor to the opt-out movement in New York. In this article, which was published in the New York Daily News, Smith writes, “Do you solemnly swear to tell the half-truth, the partial truth and nothing like the truth? Apparently, that’s the vow press officers who work for the New York State Education Department take.”

“Why was half of Thursday’s announcement of the 2014 test results devoted to testimonials, mainly from selected superintendents, on how educational gains are being made due to implementation of the Common Core?”

Smith writes that:

“The 2014 tests were given in April and scored in May, with the results and press release issued in August. Why the delay. The same pattern was followed in 2013.

“The delay is unexplained, and it feeds suspicions that the state is sitting on the results trying to figure out the most advantageous way to package them.”

Under pressure from parents, the state released half the test questions. Smith wonders, why not release them all? By the time teachers get the test results, the children are no longer in their class. Nor do teachers learn which questions their students got right or wrong. If teachers can’t learn anything about their students other than whether he or she was rated as a 1, 2, 3, o4, of what value is the test?

The state’s contract with Pearson stipulates that test questions may not be used again, so why not release to the public the entire test that the public paid for?

Sarah Garland, writing for the HECHINGER Report, says that the Reagan-era report “A Nation at Risk” (1983) laid the groundwork for today’s regime of high-takes testing, longer school hours, and tougher accountability measures. The conservative Republicans he quotes express satisfaction with the Obama administration’s embrace of their agenda. The enduring puzzle: who stole the Democratic agenda of equity and teacher professionalism?

The Vermont State Board of Education adopted a resolution on assessment and accountability with a message: We will not let the federal government bully our children. We read research and incorporate it into our policy decisions. This set of principles and resolutions could serve as a guide for every state and school district about the appropriate uses of assessment and the true goals of education in our society.

Vermont State Board of Education

Statement and Resolution on Assessment and Accountability Adopted August 19, 2014

The Vermont State Board of Education is committed to ensuring that all students develop the knowledge, capabilities and dispositions they need to thrive as citizens in their communities, higher education and their careers in the 21st century. The Board of Education’s Education Quality Standards (EQS) rules aim to ensure that all students in Vermont public schools are afforded educational opportunities that are substantially equal in quality, and enable them to achieve or exceed the standards approved by the State Board of Education.

These rules were designed to ensure continuous improvement in student performance, instruction and leadership, so that all students are able to develop high levels of skill and capability across seven essential domains: literacy, mathematics, scientific inquiry and knowledge, global citizenship, physical and health education and wellness, artistic expression, and transferable 21st century skills.

To achieve these goals, educators need to make use of diverse indicators of student learning and strengths, in order to comprehensively assess student progress and adjust their practice to continuously improve learning. They also need to document the opportunities schools provide to further the goals of equity and growth.

Uniform standardized tests, administered across all schools, are a critical tool for schools’ improvement efforts. Without some stable and valid external measure, we cannot evaluate how effective we are in our efforts to improve schools and learning. Standardized tests – along with teacher-developed assessments and student work samples — can give educators and citizens insight into the skills, knowledge and capabilities our students have developed.

What standardized tests can do that teacher developed tests cannot do is give us reliable, comparative data. We can use test scores to tell whether we are doing better over time. Of particular note, standardized tests help monitor how well we serve students with different life circumstances and challenges. When used appropriately, standardized tests are a sound and objective way to evaluate student progress.

Despite their value, there are many things tests cannot tell us. Standardized tests like the NECAP and soon, the SBAC, can tell us something about how students are doing in a limited set of narrowly defined subjects overall, as measured at a given time. However, they cannot tell us how to help students do even better. Nor can they adequately capture the strengths of all children, nor the growth that can be ascribed to individual teachers. And under high-stakes conditions, when schools feel extraordinary pressure to raise scores, even rising scores may not be a signal that students are actually learning more. At best, a standardized test is an incomplete picture of learning: without additional measures, a single test is inadequate to capture a years’ worth of learning and growth.

Along a related dimension, the American Psychological Association wrote:

“(N)o test is valid for all purposes. Indeed, tests vary in their intended uses and in their ability to provide meaningful assessments of student learning. Therefore, while the goal of using large-scale testing to measure and improve student and school system performance is laudable, it is also critical that such tests are sound, are scored properly, and are used appropriately.”

Unfortunately, the way in which standardized tests have been used under federal law as almost the single measure of school quality has resulted in the frequent misuse of these instruments across the nation.

Because of the risk of inappropriate uses of testing, the Vermont State Board of Education herewith adopts a series of guiding principles for the appropriate use of standardized tests to support continuous improvements of learning.

1. The Proper Role of Standardized Testing – The purpose of any large scale assessment must be clearly stated and the assessments must be demonstrated as scientifically and empirically valid for that purpose(s) prior to their use. This includes research and verification as to whether a student’s performance on tests is actually predictive of performance on other indicators we care about, including post-secondary success, graduation rates and future employment.

In addition, standardized test results should be used only in concert with a diverse set of measures that capture evidence of student growth and school impact across all important outcomes outlined in the Education Quality Standards.

2. Public Reporting Requirement – It is a state and local obligation to report on the quality of the schools to the citizenry. Standardized testing is part of this reporting obligation. The state board encourages local public reporting of a diverse and comprehensive set of school quality indicators in local school, faculty and community communications.

3. Judicious and Proportionate Testing – The State Board of Education advocates for reducing the amount of time spent on summative, standardized testing and encourages the federal government to reduce the current requirements for annual testing in multiple subjects in every grade, 3-8, and then again in high school. Excessive testing diverts resources and time away from learning while providing little additional value for accountability purposes.

4. Test Development Criteria – Any broad scale standardized assessment used in the state of Vermont must be developed and used appropriately in accord with the principles adopted by the American Educational Research Association, the National Council on Measurement in Education, and the American Psychological Association.

5. Value-added scores – Although the federal government is encouraging states to use value added scores for teacher, principal and school evaluations, this policy direction is not appropriate. A strong body of recent research has found that there is no valid method of calculating “value-added” scores which compare pass rates from one year to the next, nor do current value-added models adequately account for factors outside the school that influence student performance scores. Thus, other than for research or experimental purposes, this technique will not be employed in Vermont schools for any consequential purpose.

6. Mastery level or Cut-Off scores – While the federal government continues to require the use of subjectively determined, cut-off scores; employing such metrics lacks scientific foundation. The skills needed for success in society are rich and diverse. Consequently, there is no single point on a testing scale that has proven accurate in measuring the success of a school or in measuring the talents of an individual. Claims to the contrary are technically indefensible and their application would be unethical.
The use of cut-off scores reports findings only at one point on a statistical distribution. Scale scores provide significantly more information. They allow a more valid disaggregation of scores by sub-group, provide better measures of progress and provide a more comprehensive view of achievement gaps.

7. Use of cut scores and proficiency categories for reporting purposes – Under NCLB states are required to report school level test results in terms of the Percentage of Proficient Students. The federally mandated reporting method has several well-documented negative effects that compromise our ability to meaningfully examine schools’ improvement efforts:

 Interpretations based on “percent proficient” hides the full range of scores and how they have changed. Thus, underlying trends in performance are often hidden.

 The targets established for proficiency are subjectively determined and are not based on research. Interpretations based on “percent proficient” also lack predictive validity.

 Modest changes to these subjective cut scores can dramatically affect the percent of students who meet the target. Whether a cut score is set high or low arbitrarily changes the size of the achievement gap independent of the students’ learning. Thus, the results can be misleading.

So that we can more validly and meaningfully describe school- and state-level progress, the State Board of Education endorses reporting performance in terms of scale scores and standard deviations rather than percent proficient. We will comply with federal requirements, but will emphasize defensible and useful reporting metrics.

8. The Federal, State and Local Obligation for Assuring Adequacy and Equality of Opportunity – Much as the state must insure a high quality education for all children, the school must be provided with adequate and equitable resources from the federal, state and local governments and must use these resources wisely and judiciously. Thus, any report on a school based on the state’s EQS standards must also include a report on the adequacy of resources provided by or to that school in light of the school’s unique needs. Such evaluations shall address the adequacy of resources, the judicious use of resources and identify any deficiencies.

Resolution on Assessment and Accountability Vermont State Board of Education

WHEREAS, our nation and Vermont’s future well-being relies on a high-quality public education system that prepares all students for college, careers, citizenship and lifelong learning, and strengthens the nation’s and the state’s social and economic well-being; and

WHEREAS, our nation’s school systems have been spending growing amounts of time, money and energy on high-stakes standardized testing, in which student performance on standardized tests is used to make major decisions affecting individual students, educators and schools; and

WHEREAS, the overreliance on high-stakes standardized testing in state and federal accountability systems is undermining educational quality and equity in the nation’s public schools by hampering educators’ efforts to focus on the broad range of learning experiences that promote the innovation, creativity, problem solving, collaboration, communication, critical thinking and deep subject-matter knowledge that will allow students to thrive in a democracy and an increasingly global society and economy; and

WHEREAS, it is widely recognized that standardized testing is an inadequate and often unreliable measure of both student learning and educator effectiveness; and

WHEREAS, a compelling body of national research shows the over-emphasis on standardized testing has caused considerable collateral damage in areas such as narrowing the curriculum, teaching to the test, reducing love of learning, pushing students out of school, and undermining school climate; and

WHEREAS, high-stakes standardized testing has negative effects for students from all backgrounds, and especially for low-income students, English language learners, children of color, and those with disabilities; and

WHEREAS, the culture and structure of the systems in which students learn must change in order to foster engaging school experiences that promote joy in learning, depth of thought and breadth of knowledge for students; therefore be it

RESOLVED that the Vermont State Board of Education requests that the Secretary of Education reexamine public school accountability systems in this state, and develop a system based on multiple forms of assessment which has at its center qualitative assessments, does not require extensive standardized testing, more accurately reflects the broad range of student learning, decreases the role of compliance monitoring, and is used to support students and improve schools; and

RESOLVED, that the Vermont State Board of Education calls on the United States Congress and Administration to accordingly amend the Elementary and Secondary Education Act (currently known as the “No Child Left Behind Act”) to reduce the testing mandates, promote multiple forms of evidence of student learning and school quality, eschew the use of student test scores in evaluating educators, and allow flexibility that reflects the unique circumstances of all states; and

RESOLVED that the Vermont State Board of Education calls on other state and national organizations to act in concert with these goals to improve and broaden educational goals, provide adequate resources, and ensure a high quality education for all children of the state and the nation.

I apologize to you, dear readers, in advance, but I must ask you to read the latest balderdash written by someone who works for Secretary of Education Arne Duncan. From my days working in the U.S. Department of Education in 1991-93, I know full well that Cabinet Secretaries have several writers and don’t actually write anything themselves. Okay, so this latest statement from Duncan says that there is too much emphasis on testing. Testing is taking the joy out of teaching. It is sucking the oxygen out of the nation’s classrooms. Nowhere does he acknowledge that his very own Race to the Top demanded more high-stakes testing, demanded that teachers’ evaluations depend on the test scores of their students. Nowhere does he acknowledge his cheerleading for VAM–value-added measurement–or his hearty congratulations to the Los Angeles Times when it published the ratings of teachers based on the test scores of their students. Over the past five years, we have learned that what Arne says bears little relation to what he does. In the same breath, as this statement shows, he is both for and against testing. He seems not to see the connection between toxic testing and the policies he has put in place.

Fortunately, two of our best thinkers have written excellent responses to the new Duncan line on testing.

Anthony Cody says that Duncan is responding to the call of Gates for a moratorium (the point is illustrated by an old advertisement for a phonograph that said “his master’s voice”). He also believes the new tack is Duncan’s response to polls that show a decline in support for the Common Core. Cody points out that the most onerous demands for high-stakes testing were initiated by Arne Duncan. What is Duncan really offering, asks Cody: a one-year moratorium on the punishments attached to testing.

Cody writes:

“But a one year deferral does not do much to fundamentally alter the systemic change that is under way. The new Common Core tests are still being rolled out and will be given this coming spring. This only amounts to a one year delay to the time when those scores will be used for evaluative purposes.

“Duncan makes it clear that the purpose of this delay is to allow for a successful transition to the new standards, testing and evaluation systems. There is actually no real change in any of the substance of any of these programs, and he reiterates the Department’s commitment to the new tests.

“If Duncan is serious in his concern about tests are “sucking the oxygen” out of schools, he should begin to listen to teachers when they tell him to stop using these tests for their evaluations and to close schools. Until then, test scores will continue to rob children of the vital learning environments they need, and teachers will continue to object.”

Peter Greene also has a withering analysis of Duncan’s new line on testing.

Greene writes:

“Duncan is shocked– shocked!!– that anyone would think it’s a good idea to make a high stakes test the measure of student achievement or teacher effectiveness. “Growth is what matters. No teacher or school should be judged on any one test, or tests alone –” And here comes the vertiginous woozies (dibs on this as a band name) again, because that would be a heartening quote if it did not come from the very same office which decreed that by order of the federal government high stakes tests must be used as a measure of student achievement and teacher effectiveness. Duncan is talking about this test-based evaluation of students and teachers as if it just spontaneously occurred, like some sort of weird virus suddenly passed around at state ed department sleepover camp, and not a rule that Duncan’s office demanded everyone follow. Has Duncan forgotten that he just made the entire state of Washington declare itself a Failing School Disaster Zone precisely because they refused to use high stakes tests as a measure of student achievement and teacher effectiveness?”

And Greene adds:

“As far as Duncan’s other concerns go– a year will not matter. Much of what he decries is the direct result of making the stakes of these tests extremely high. Student success, teacher careers, school existence all ride on The Test. As long as they do, it is absurd to imagine that The Test will not dominate the school landscape. And that domination is only made worse by the many VAMtastic faux formulas in circulation.

[Says Duncan: Too much testing can rob school buildings of joy, and cause unnecessary stress. This issue is a priority for us, and we’ll continue to work throughout the fall on efforts to cut back on over-testing.]

Oh, the woozies. Duncan’s office needs to do one thing, and one thing only– remove the huge stakes from The Test. Don’t use it to judge students, don’t use it to judge teachers, don’t use it to judge schools and districts. It’s that attachment of huge stakes– not any innate qualities of The Test itself– that has created the test-drive joy-sucking school-deadening culture that Duncan both creates and criticizes. If the department doesn’t address tat, it will not matter whether we wait one year or ten– the results will be the same.

Give it up, reformers. The scores on the ACT are flat from 2010-2014, despite the billions wasted on testing, test-based teacher evaluation, and merit pay. Your reforms have reformed nothing. They have failed. Pay attention.

Improve the lives of children and families. Improve working conditions in the school. Demand equitable resources for schools. Reduce class sizes for needy children. Do what works. Throw your punishments and sanctions into the ash-heap of history. It will happen sooner or later.

Start now to build the structures that work for students and teachers.

FairTest_______________________
National Center for Fair & Open Testing
for further information:
Bob Schaeffer (239) 395-6773
cell (239) 699-0468
for use with annual ACT scores on or after Wednesday, August 20, 2014

STAGNANT ACT SCORES SHOW TEST-DRIVEN U.S. SCHOOL POLICIES
HAVE NOT IMPROVED COLLEGE READINESS,
EVEN WHEN MEASURED BY OTHER TESTS

Another year of flat scores on the ACT, the nation’s most widely administered college admissions exam, provides further evidence that a decade of test-driven public school policies has not improved educational quality.
Reacting to ACT scores released today, Bob Schaeffer, Public Education Director of the National Center for Fair & Open Testing (FairTest) said, “Proponents of ‘No Child Left Behind,’ ‘Race to the Top,’ ‘waivers,’ and similar state-level programs promised that focusing on testing would boost college readiness while narrowing score gaps between racial groups. The data show a total failure according to their own measures. Doubling down on unsuccessful policies with more high-stakes,
K-12 testing, as Common Core exam proponents propose, is an exercise in stubbornness, not meaningful school improvement.” (see http://fairtest.org/common-core-assessments-factsheet)

Stagnant scores and racial gaps have also been reported on the federal government’s National Assessment of Educational Progress (NAEP) and the SAT college admissions test.

Schaeffer continued, “The lack of progress toward excellence and equity will provide further ammunition for the country’s growing testing resistance and reform movement. Ending the counter-productive fixation on standardized exams is necessary to create the space for better assessments that actually enhance learning and teaching.” FairTest actively supported this past spring’s opt-out campaigns and other protests that focused attention on testing overuse and misuse.

FairTest is also a national leader for test-optional higher education admissions. More than 830 accredited, bachelor-degree granting colleges and universities now do not require all or many applicants to submit SAT or ACT scores (see http://fairtest.org/university/optional). Eight more schools – Wesleyan University, Old Dominion University, Hofstra University, Temple University, Montclair State University, Beloit College, Bryn Mawr College and Emmanuel College — dropped test-score requirements already this summer. In addition, Hampshire College, which long was test-optional, is now “test-blind.”
– – 3 0 – -

2014 COLLEGE-BOUND SENIORS AVERAGE ACT SCORES
1,845,787 million test takers

COMPOSITE SCORE FIVE-YEAR SCORE TREND
(2010 – 2014)
ALL TEST-TAKERS 21.0 0.0

African-American 17.0 + 0.1
American Indian 18.0 – 1.0
Asian 23.5 + 0.1
Hispanic 18.8 + 0.2
White 22.3 0.0

Source: ACT, The Condition of College & Career Readiness 2014

The first results of Utah’s Common Core tests are in, and they follow the pattern of other states: a sharp drop in the proportion of students who are “proficient.”

“The percentage of Utah students who scored proficient or better in science ranged from 37 percent to 45 percent, depending on grade level. In math, anywhere from 29 percent to 47 percent of kids scored proficient. And in language arts, proficiency ranged from 38 percent to 44 percent.

“Proficiency was defined as performing at or above standards for grade level.”

“Proficiency levels were much higher on CRTs last year. In science last year on CRTs, proficiency levels ranged from 58 percent to 76 percent, depending on grade level; in math, from 39 percent to 85 percent; and in language arts, from 77 percent to 90 percent.”

State officials, having bought into the Common Core, are not at all disturbed.

Parents and teachers should be outraged. Common Core tests are aligned with NAEP (National Assessment of Educational Progress) “proficiency” levels, which are NOT grade level. NAEP proficiency represents “solid academic achievement.” I served on the NAEP governing board for seven years. Most students will not reach NAEP proficient because it is NOT grade level. I think of it as a high level of achievement.

May I remind you that we are one of the most powerful and most creative and most productive nations in the world. We didn’t get that way with a stupid population.

The Common Core tests are developmentally inappropriate. The achievement levels are out of reach of most students. You cannot reasonably expect a fifth-grader to answer questions on a seventh-grade level? Does Utah have plans for the 50-55% of their students who will not be able to graduate high school because of Common Core’s absurd definition of “proficient”?

Florida has gone bonkers. State law requires children in kindergarten to take tests for every subject taught in kindergarten. Some counties will develop as many as 15 different tests, ranging from physical education to art. Most children will be required to take seven tests.

State Sen. David Simmons, a member of the education committee, said “For us to assure that schools do their jobs we can only test. If you don’t test, you don’t care,” said Simmons.

School districts will decide whether children’s performance on the tests will impact their grades and their ability to move on to first grade.

This must be another of former Governor Jeb Bush’s bright ideas. You can tell how much he cares because Florida students take so many tests.

On Anthony Cody’s new independent blog site, “Living in Dialogue,” Chicago teacher Michelle Gunderson offers her views on the ethical use of student data. 

 

In her many years as an elementary school teacher, she has seen standardized tests evolve from a sorting instrument to a means of punishing children to an excuse for privatizing public schools.

 

She will not be complicit in any of these uses of student test scores. She would abolish the standardized tests if she could, but that is not within her power.

 

So she pledges, first, that they will always be on of multiple measures; that she will remain strict confidentiality about student test scores and never publish them on a data wall or release them to the public; and that she will communicate with families about the frequency and amount of time spent on testing.

 

Tests, like all tools, may be used wisely or wrongly. Tests should be used to help children and teachers, not to punish or label them or close their school.

Carol Burris and Biana Tanis take a close look at New York’s Common Core tests and find them deeply flawed. Burris is a high school principal on Long Island and Tanis is a public school parent and special education teacher in the Hudson Valley.

State officials celebrated paltry results: the passing rates on the reading test were flat and increased in math by 4.6%. But nearly 2/3 of the state’s children did not reach the state’s unreasonably high proficiency level. Testing experts and state officials knew in advance what the results would be. Why do they stubbornly cling to the outworn cliche that raising the bar improves achievement? We thought that idea was discredited by the abject failure of NCLB. If a run er can’t clear a four-foot bar, how will he clear a six-foot bar?

Burris and Tanis show that certain groups, such as students with disabilities and English language learners, did very poorly. The content was far beyond their capacity. Fifty percent of the questions were released, and the authors show that many were age-inappropriate. What is the logic of giving 7th grade content to a 5th grader?

Here is an example:

“In addition to passage difficulty, the questions themselves required skills out of the reach for many young children. Consider this fourth-grade question on the test based on a passage from Pecos Bill Captures the Pacing White Mustang by Leigh Peck.

Why is Pecos Bill’s conversation with the cowboys important to the story?

A) It predicts the action in paragraph 4

B) It predicts the action in paragraph 5

C) It predicts the choice in paragraph 10

D) It predicts the choice in paragraph 11

Visualize the steps required to answer this question. First, 9-year-olds must flip back to the conversation and re-read it. Next, they must go back to the question and then flip back to paragraph 4. Complete this step 3 more times, each time remembering the original question. In addition to remembering the content of each paragraph, they must also be mindful that choices A and B refer to the action in the related paragraph, while choices C and D refer to a choice. Similar questions were on the third-grade test. Questions such as these are better suited to assess one’s ability to put together a chair from Ikea than they are to assess student’s understanding of what they read.”

Is it any wonder that parent anger towards the Common Core is growing in New York? This is a blue state; these parents are outraged by a state policy that labels their children as failures based on tests that are developmentally inappropriate. Why does the state want 2/3 of its children to be branded as failures? If this is what Common Core means, it will have a short life indeed. It may be fine for the kids bound for the Ivy League, but most kids are not. We need common sense more than Common Core.

There is a secret society that was created by a couple of dads who are teachers on Long Island. They were worried about all the testing and the way that schools were misusing the test results to label kids.

These dads wanted to protect their children–their own and the ones they teach and even the ones they don’t teach–from practices that they knew were harmful.

But what to do?

First they dressed up in funny costumes, but that didn’t get them far.

Eventually they settled on the idea of green laces, either shoelaces or wristbands.

They called their group “Lace to the Top,” to symbolize kindness and concern for children, as opposed to the aggressive competition encouraged by Race to the Top.

If you watch their video, you might catch a glimpse on me being interviewed by Jon Stewart on the Daily Show. I was wearing a green bracelet; I gave the laces to Jon Stewart. I hope he wears them.

This is a gentle kind of resistance, but it is resistance nonetheless.

Its message: Our children are more than a test score. We care about all kids, even those who don’t get the top test scores. After all, only a few kids get the top test scores, and what kind of society throws away the majority of its children because they are not at “the top?” Every child is precious and deserves the best education we can give them and many opportunities to develop into good people.

 

Is the message of “Lace to the Top” too gentle to overcome the “survival of the fittest” mentality that is implied by Race to the Top? Funny how history repeats itself. In school, we learned about social Darwinism, and we thought it was an obsolete ideology. We were taught that our society had long ago outgrown the philosophy that the strong win and the weak die off. We thought that our society had learned some lessons about social responsibility. And yet, here we are, more than a century later, with a federal policy that explicitly encourages survival of the fittest.

 

The dads who created Lace to the Top want us to think of all children as if they were our own. No winners or losers. Children. Each of them deserves the best education that we can provide and that they need. If you agree, get your green shoelaces and Lace to the Top.

Follow

Get every new post delivered to your Inbox.

Join 108,895 other followers