The Fraud of High Stakes Testing and “Accountability”

To fully understand the pernicious agenda of “education reform” we need to look more closely at the reform “strategies”. First of all there is the idea of “accountability”. Who could disagree that teachers and schools should be held accountable for what they do? Accountability is a great slogan but the reality of what has been done under the cover of this slogan is a giant fraud perpetrated on students, parents and teachers.

The key measure promoted to assess the performance of students, teachers and schools by the reformers is “high stakes” tests. Socialists are not opposed to testing per se although we would radically scale back its use. Along with other means of assessment, standardized tests can, with many limitations, help to identify areas where students are weak and need further assistance. On the other hand, there are many things taught by good teachers which are very hard to assess using a multiple choice test.

High stakes testing, however, goes much further than simply using testing to help drive instruction. It is the dogmatic belief that data from a single reading and math test given once a year should be used as the basis of all sorts of decisions including the closing of “failing” schools as well as the evaluation and payment of teachers. Reformers further argue that teachers should be fired if their “performance”, i.e. their students’ test scores are persistently low. Layoffs likewise should not be based on seniority but on test-determined “performance”.

The reformers furthermore completely ignore the radical differences in the challenges faced by schools in different communities. There is no understanding that schools teaching large numbers of vulnerable or marginalized children could have successes measured on other levels than standardized tests. How does the high stakes approach measure the role a school plays in supporting its community, helping teenage mothers stay in education, dealing with children in homeless shelters or moving from one foster home to another? How does it take into account the problems facing schools with large numbers of English language learners?

In 2001, No Child Left Behind enshrined high stakes testing as the key tool of federal education policy. Federal education funding was made contingent on states developing and administering high stakes tests in all grades from three to eight and determining which schools were “failing” based on formulas for “adequate yearly progress” on tests. Parents were to be given the option of taking their children out of failing schools which would be shut if they did not “turn around”.

NCLB put forward the bizarre and manifestly unachievable goal that all students would be “on grade level” by 2014. This seemed to sum up NCLB: it was a set up for failure and the failure would justify further, deeper attacks on public education. This is of course exactly what has happened.

While NCLB is remembered as George Bush’s policy it had strong bipartisan support. A co-sponsor of the bill in Congress was the late liberal Democratic Senator Ted Kennedy. As the act’s name suggests there was a lot of lofty rhetoric at the time from Bush and others about the “bigotry of low expectations” that supposedly permeated education and how high stakes testing and accountability were going to ensure the end of substandard education for inner city black youth.

Some, including ourselves, opposed high stakes testing from the start for a range of reasons. But nine years after the passing of NCLB it is possible to state categorically that none of the claims made has been delivered upon.

First of all high stakes testing has predictably led to pervasive “teaching to the test” as schools and individual teachers desperately try to keep their heads above water. The whole school year is reduced to performance on a single literacy and math test. Especially in inner city schools this has led to the downgrading of social studies, science, critical thinking or anything that doesn’t directly help get the numbers up. The most tragic result is the cuts in art and music programs. In this way, even the pretense of a rounded education has been abandoned.

But the pernicious effects go deeper. Even in the “core” areas of literacy and math focused on by high stakes testing the result of teaching to the test is to narrow the curriculum and thus degrade the quality of instruction. This is greatly facilitated by the publishing of previous years’ tests which are excellent predictors of what will come up on future exams. Teachers are forced to focus on the “skills” which are expected to come up as well as on test-taking itself. Incessant “test prep” may lead to improved scores on the state tests but they do not indicate a broad understanding of the grade level material.

This is clearly shown when students are given different literacy and math tests for which they have not been “prepped”. The federal government administers the National Assessment of Educational Progress (NAEP) to a portion of the student population around the country. Since these are not “high stakes” tests the data is a much more reliable measure of student understanding than the state tests. The most recent data from the NAEP shows that in the eighth grade nationally there have been no gains in reading for the past 10 years. In math, the gains were more substantial before NCLB.

So at the most basic level, the claim that high stakes tests will tell us about the real level of student knowledge and “growth” is laughable. What these test scores do measure is wealth. Study after study has found a very strong correlation between high tests scores and the affluence of the neighborhoods where the students come from. As Karen Lewis, the recently elected president of the Chicago Teachers Union puts it, “high-stakes testing basically tell us more about a student’s socioeconomic status than it does anything else.”

What high stakes test scores do is penalize students in poorer neighborhoods, by deeming their schools “failing” This leads to a paradox where the schools that need the most help and resources will actually get less and less help from the state, thus leading to the eventual collapse of these schools.

This is a fundamental tactic of conservative ideology: starvation. Students from low-income, racial minority, or single-parent families are often strained by low wages, poverty, long working hours, and discrimination. These students have to struggle to overcome many more obstacles than students from middle-class, wealthy, or white families. The government can impose all the tests on schools that they want, but until we have a government that is willing to tackle these central problems, student achievement from working-class communities and communities of color will generally continue to lag behind.

Merit Pay

Since these tests essentially measure nothing important related to education, how can they be used to decide which schools stay open, how teachers are paid, who gets tenure, etc? But the education deformers are nothing if not brazen and shameless.

Take merit pay. Socialists oppose merit pay because it promotes division among education workers with no tangible benefit to students. Given that it is based solely or mainly on test scores, merit pay means even more teaching to the test. It also means the work of teachers in very different situations is reduced to one highly dubious quantitative measurement. Ultimately, it is a step towards firing teachers whose students did not perform well enough on tests.

In a particularly outrageous but logical example of where this can lead the Los Angeles Times recently published a searchable database of 6,000 Los Angeles teachers which purported to ranks them from least to most “effective” based on the “value added” by these teachers to their students. The “value added” is measured of course through high stakes test scores, comparing how the given students did at the start of the school year with the end of the year.

More and more school districts are using “value added” measures as part of teacher evaluations or in the case of Washington DC as a basis for firing those deemed “poorly performing”. It is claimed that this method of measuring teacher performance is fairer than the NCLB method of comparing succeeding classes, i.e. this year’s fifth graders versus last year’s fifth graders. However, the problems with the “value added” method of measuring teacher performance are numerous to say the least. For example: many teachers could theoretically contribute to a child’s improved scores on a literacy or math test. How can we determine statistically the amount played by each? In a given year many students move school. Mobility is especially high in poor neighborhoods. Should students that a teacher has taught for a month be part of his or her evaluation?

One could go on like this but it is enough to quote Edward Haertel, a Stanford professor who wrote a report on “value added” methods: “If these teachers were measured in a different year, or a different model were used, the ranking might bounce around quite a bit…people are going to treat these scores as if they were reflections on the effectiveness of the teachers without any appreciation of how unstable they are.” (New York Times, 9/1/10) Initially Arne Duncan hailed the LA Times for publishing the list of teachers but he has since backed off a bit. In fact the Department of Education’s own research wing did a study on “value added” methods which concluded that they “are subject to a considerable degree of random error.”

How New York City’s gains magically disappeared

But to see the full dimensions of the high stakes testing fraud one must look more closely at how it has operated in practice in specific school districts. Take the example of New York City where Michael Bloomberg made education his signature issue when he became mayor in 2002. He abolished the old Board of Education and established direct “mayoral control” of the school system, following Chicago’s example.

Since then he and his schools chancellor Joel Klein have presided over eight years of intensive education “reforms” which have been hailed by the corporate media around the nation. Endless accounts have been written of the “gains” in literacy and math especially by black and Latino students in the city’s poor neighborhoods. This was based on the spectacular rise in students scoring “at grade level” or “above grade level” on the state tests. In 2009, 82 percent of city students passed the statewide test math and 69 percent in English, up sharply from 42 and 38 percent, respectively, in 2002. During the same period, the so called “racial achievement gap” in passing rates was cut in half on some tests.

How was this amazing progress achieved? The key to answering this question is that the New York State Education Department was in charge of creating the tests and setting the scores required for passing. This is true around the country and it gives the states a tremendous incentive to “dumb down” the tests in order to make their numbers look better and thereby reduce the number of schools in their area deemed to be failing. This is a point often made by none other than Arne Duncan. And because so much is at stake there is tremendous temptation for individual teachers, schools and whole school districts to engage in systematic, old fashioned cheating. Such schemes are being regularly exposed.

In New York State the method used to game the results was not primarily to make the tests easier. Rather the state education authorities dramatically lowered the score required to pass. It reached the point where even some sections of the normally cheerleading media began to ask questions. The New York Times (9/14/09), for example, did a study which showed that on the sixth grade English test a student who randomly guessed had an 89 percent chance of getting 44 percent right which would earn him or her a “somewhat behind grade level” score sufficient to pass to the next grade.

In July 2010 the house of cards came crashing down. New York State raised its passing scores. Seemingly overnight most of the academic “gains” made by city students, as measured by high stakes tests, simply evaporated into thin air. The passing rate went from 82 percent in math and 69 percent in English to 54 percent and 43 percent respectively, losing the bulk of the false “gains” of recent years. And given that the federal NAEP test shows a mere 22 percent of eighth graders in the city are “proficient” in reading, the new state scores are likely still a serious exaggeration.

Schools in poor neighborhoods fared even worse. For example the drop in passing scores on the math tests was 35 points for black k-12 students (75 percent to 40 percent) compared to a 17 percent drop for white students (92 percent to 75 percent). This is because the claim that increasingly large numbers of black and Latino students were on or above grade level masked the reality that a large portion of these students’ scores were just above the “grade level” cutoff. In middle class schools, students scores were generally well above the cutoff. So not only did the overall gains claimed in the city disappear but the “racial achievement gap” that was supposedly closing suddenly reopened as large as ever.

What was the reaction of Bloomberg and Klein? Bloomberg declared “This doesn’t mean the kids did any worse – quite the contrary. What this is simply saying is that we’ve redefined what our objectives are for the kids. Whether the new expectations will instigate all of us to try harder, one can only hope.” The shamelessness of this is difficult to convey. Bloomberg is one of the richest men in the world who built his fortune based on selling financial data. He has made “data” the driving force of his business oriented overhaul of education. Teachers are ceaselessly told in the media that in the private sector people who are incompetent or fail to achieve results are fired. But after the empire of data was exposed as an empire of lies is there any sign of Schools Chancellor Klein resigning? Or even an apology? Not a bit of it. In the private sector people sometimes actually go to jail for fraud on this scale.

A disservice to students and their parents

One of the aspects of high stakes testing that is not much discussed is the absolute disservice that has been rendered to so many working class students and their parents who have been told that they are “on grade level” in reading or math when in reality many are several years or more behind. It demonstrates how the talk of providing quality education for all children regardless of their background is meaningless. As we said earlier, the corporate elite could care less whether the mass of working class youth are really getting a decent education as long as they get a docile workforce with a basic skills set.

And just in case any doubts remain on the question, teaching to the test and endless test prep is not how the wealthy want their children educated. This “model” of education is simply not to be found in the better sort of private schools. But it’s good enough apparently for everyone else and especially for the children of the poor.

For these students what lesson does high stakes testing really instill? It shows that what society, i.e. the ruling class, values is not critical thinking and open ended enquiry or even knowledge itself but rather the mastery of a set of procedures. Preparing for literacy tests, students are taught to figure out what the “test writer” is looking for in each question regardless of whether the student has any interest in or overall comprehension of the material. Endless testing and test prep can be seen as a form of job training though not for any sort of job that young people would really aspire to have. The key thing is not understanding of the material but following directions mindlessly. But when the ruling class can’t even provide low-paying, dead-end jobs for them, young people will increasingly reject the system which thinks high stakes tests are a good preparation for life.