By Joseph Herbert
You’ve heard the narrative before. Our schools are broken. Our kids don’t know anything; they’re not ready for 21st century jobs. We have terrible teachers. Education is the civil rights issue of our time. We need to fix the status quo.
This narrative paints a bleak picture, and many well-intentioned reformers have devised “solutions” to “fix” education in America. What has been left out of the narrative, however, are the unintended and harmful effects of education reform, starting with No Child Left Behind (NCLB) and continuing unabated with Race to the Top.
One of the core tenets of the new, reform-based status quo is an over-reliance on testing. The argument is enticingly straightforward: give a test to see where the good and bad teachers and schools are, then reward the good and punish, fire, and close the bad. What could possibly be the problem?
The problem is two-fold: (1) the data collected do not reliably give us the information that reformers claim they do, and (2) the over-emphasis on testing is sucking the life and joy out of school, interfering significantly with actual teaching and learning, and narrowing the curriculum.
Problem 1: Data Quality
Let’s begin with the first problem: the quality of the data produced by standardized tests.
Like all states that receive Race to the Top money and NCLB waivers, D.C. Public Schools (DCPS) evaluates many of its teachers according to a value-added measurement (VAM) model, with the exception of a one-year moratorium this past school year. The idea behind VAM is to quantify how much one specific teacher added to students’ learning, while controlling for other variables such as prior knowledge and student background.
The American Statistical Association has strongly cautioned against using VAM models in high-stakes decisions, noting that, “VAMs typically measure correlation, not causation: Effects – positive or negative – attributed to a teacher may actually be caused by other factors that are not captured in the model.” Furthermore, “Most VAM studies find that teachers account for about 1% to 14% of the variability in test scores, and that the majority of opportunities for quality improvement are found in the system-level conditions. Ranking teachers by their VAM scores can have unintended consequences that reduce quality.”
VAM models do have broad, diagnostic value when their use is strictly tempered with an understanding of their limitations, but that is not how they are currently used in education decision-making. While reformers do sometimes pay lip-service to the need to include other measures of teacher effectiveness, their mantra has been for VAM scores to become increasingly important as a supposedly pure and perfect reflection of a teacher’s or school’s effectiveness. Simply put, education reformers want to use standardized test data to determine how effective teachers are in a way that is not statistically valid.
Problem 2: Lost Instructional Time
As a math teacher at Wilson High School in Washington D.C., I am incredibly concerned about the amount of instructional time my students lost this year to the new Common Core PARCC tests.
Our freshmen and sophomores took over seven hours worth of PARCC tests in the month of March alone. Furthermore, they had a second round of PARCC tests in May followed shortly thereafter by final exams in June. Ultimately, these children had three out of the last four months of school dominated by tests.
Previously, all students took the paper-and-pencil DC-CAS standardized test at the same time, and instruction was disrupted for about a week. With the new PARCC test, there was a much more protracted disruption to instruction. The PARCC test is administered online, but Wilson simply does not have the technological infrastructure to test large numbers of students simultaneously. Without the necessary IT infrastructure we were forced to test small groups of students on a rotating basis.
As a result, we spent over three weeks administering the first round of PARCC tests alone. Students were forced to miss class to test while their classes went on, causing them to lose valuable instructional time.
Furthermore, some classes had 1,125 instructional minutes during PARCC testing, while others had only 512 as we adjusted our schedule to accommodate the testing times. This meant that students spent less than half as much time learning some subjects as they did others.
Ultimately, PARCC testing caused significant and severe disruption to all teaching and learning at Wilson High School, even affecting older students who did not take a single PARCC test.
I want to be clear: I am not blaming Wilson’s administration for this outrageously disruptive schedule. They did the best that they could, given the testing mandates from downtown and our lack of infrastructure and resources.
I also don’t want to be accused of blaming the Common Core State Standards. There has been vocal criticism of the Common Core’s Language Arts standards, but as a math teacher, I can only speak to the math standards. I love the Common Core. It is far from perfect, but it represents a vast improvement over D.C.’s old math standards. The issue is not the Common Core. The issue is that we are forced to give tests like PARCC.
I’ve heard DCPS representatives say many times that they are aware that there is too much testing. They say they are working to make things better, but the PARCC tests make this problem exponentially worse.
While it is true that PARCC has announced some consolidation of its tests for the coming school year, if the PARCC tests are implemented as planned next year, they will yet again steal valuable instructional time from students, create massive disruptions to schools across the city, and fail to provide educators with actionable data to improve their classroom instruction.
Alternatives and a Path Forward
As the NCLB rewrite is being debated in Congress, many of its supporters say that if we abolish annual testing, we won’t have any data showing us that the achievement gap exists. This is simply false. The National Assessment of Educational Progress (NAEP) uses statistical sampling to generate valuable data on student achievement across the country. It does this without obtrusively turning schools into testing prisons and stealing time from valuable and meaningful instruction. And unlike the brand new PARCC, NAEP provides longitudinal data dating back to the 1970s.
Ultimately, we have to ask ourselves what kind of experience we want our children to have in school. I believe that kids learn more when they’re excited to come to school. I believe they learn more when they have meaningful and thoughtful questions to ask and answer. I believe math can and should be fun.
I believe that if we want to stem the tide of DC’s dropout crisis, school should be a worthwhile place to attend, not a miserable experience of test taking.
I believe that if we want to close the achievement gap, we need to have an open and honest conversation about what students need from school, what they want from school, and how we can get data on student performance without perversely affecting their school experience.
Most of all, I believe that school is a place for profound growth and learning. Anything that detracts from or actively impedes that must go.