AFT Analysis: K-12 Testing Is Pervasive, Costly
Tests and test preparation cost hundreds of dollars and weeks of instructional time in America's classrooms—the equivalent of an entire class period a day in one test-heavy district, contends a report released today by the American Federation of Teachers.
The analysis, based on a deep dive into two districts' testing programs, analyzes the expenditure of both time and money associated with three different categories of tests: the annual summative assessments in grades 3-8 required by the No Child Left Behind Act and state law; "interim" assessments used to predict how students will fare on the state tests; and local tests ranging from diagnostic reading assessments to curriculum exams crafted by teachers.
In all, the report concludes, states and districts should take a fresh look at their entire testing apparatus. They should gather feedback from teachers in order to cut tests that are either duplicative or unhelpful in planning and delivering instruction. The savings, the report argues, could be invested in better-quality tests and used to restore instructional time, including for art, music, and PE programs.
Obliquely, the report also underscores that many testing decisions are under the purview of districts, rather than states or the federal government, where the NCLB law is often faulted for expanding testing.
The AFT has ramped up its criticism of tests in recent years, so the findings should be considered through that lens. Still, they offer a fresh look at an area of K-12 policymaking in which anecdotes are rampant and hard data are difficult to find.
For the report, AFT researcher Howard Nelson gathered the testing calendar and assessment inventory from two unidentified medium-sized urban districts, one in the Midwest and one in the East.
To estimate the costs of the tests, he drew on recent research of the costs of state and interim tests, and coupled it with the direct costs of local assessments based on district reports. And he drew on additional research studies to approximate the amount of time spent on test preparation.
Among the findings:
- The tests themselves took up about three days of time, annually, in the Midwest district and two full weeks annually in the Eastern district.
- Annual preparation for tests, narrowly defined as taking practice tests or drilling in test-taking strategies, took up an additional 16 days in the Midwest, and a month in the second district.
- Depending on grade level, tests cost, including both test administration and lost instructional time, was as high as $600 per student in the Midwest district and as high as $1,100 per student in the Eastern district.
- The districts varied in which grades were the most heavily tested and the most costly, as well as what types of assessments they emphasized.
Overall, the report indicates that there is an awful lot of testing going on, though the precise amounts vary district by district.
"I can't say the findings are generalizable or scientific yet, but they give a reasonable range of the amount of testing," Nelson said.
Although the report attributes the rise in testing in part to the NCLB law, it shows that much testing time is spent on assessments that are not federally required. In the Midwestern district, for instance, state testing to meet NCLB requirements took up 390 minutes in grades 3-8 overall. But that figure is dwarfed by the 930 minutes spent on ACUITY, a CTB McGraw-Hill interim-test product, which is not required under federal law.
Even so, accountability pressures and new incentives, such as the teacher-evaluation requirements of the Race to the Top program, appear to have taken their toll, Nelson said.
"What happened is once you're held accountable in a really serious way for state-mandated tests, everyone starts preparing for them, because they don't know what else to do," he said.
The report is highly critical of the interim assessments, portraying them as low-quality and recommending that districts and states eliminate them. "Most educators and many experts believe that interim/benchmark tests are not useful for improving instruction," it says, citing research that supports that conclusion.
As for its recommendations that districts pare back their test schedules, Nelson acknowledged that teachers hold very different opinions about the value of assessments, so figuring out which ones to scrap and which ones to preserve won't necessarily be an easy task. The Eastern districts' curriculum-based assessments, which are developed and scored by teachers, take up a good chunk of time but tend to be popular, he said.
"All year long as I worked on this, I got entirely different responses from teachers," he said. "Some teachers will say DIBELS [a reading-skills inventory] is worthless, and others say they like it. Maybe districts should offer a menu of tests schools could select from, rather than mandating them to be the same in every school in the district."