The Washington PostDemocracy Dies in Darkness

Beyond test scores: The right way to assess students and schools

Placeholder while article actions load

For years students, schools, teachers, principals and even states have been “graded” by student standardized test scores under the false premise that the scores alone are legitimate measures of how much a student has learned, how well a teacher has taught, and how schools are closing the achievement gap.

Now that the test-centric No Child Left Behind law has been retired in favor of the new Every Student Succeeds Act, there is a new opportunity for states to construct new assessment systems that make more sense, and a new report details how such a system could be constructed.

The report, “Assessment Matters: Constructing Model States Systems to Replace Testing Overkill,” was just issued by the National Center for Fair and Open Testing, or FairTest, a non-profit dedicated to ending the abuse and misuse of standardized tests. This post is based on the report, and it was written by Monty Neill, executive director of FairTest.

By Monty Neill

The way we measure students’ academic progress sends powerful messages about what kinds of learning we value. When measurement systems are used to evaluate schools, the factors they emphasize can control classroom practices, for good or ill.

The test-and-punish approach embodied in the federal No Child Left Behind (NCLB) law undermined educational quality for many. It inhibited school improvement. It delivered a message that deep learning and supportive, healthy school environments do not matter.

The damage has been most severe in the most under-resourced communities. There, the fixation on boosting test scores not only harmed teaching and learning, it also led to mass firings and school closings. The deteriorating educational climates fed the school-to-prison pipeline.

The Every Student Succeeds Act (ESSA), which replaced NCLB, allows states to shift the thrust of accountability away from punishing schools and teachers. Instead it allows them to focus on providing genuine help to improve educational quality and equity.

ESSA also includes an “Innovative Assessment” pilot project, which opens the door to significantly better assessments. The word “assess” comes from the Latin term meaning “to sit beside.” Assessing implies a direct and active relationship between or among people. Assessing could involve an observation by a teacher, a conversation between teacher and student, or reviewing a student’s work. It typically includes interaction to provide feedback or find out more about the student’s thinking or depth of knowledge.

A new FairTest report guides states and districts in moving toward teacher-led, student-focused, classroom-based performance assessments. Assessment Matters: Constructing Model State Systems to Replace Testing Overkill describes a system that could be built under ESSA. Its primary purpose is to support high-quality, truly personalized student learning. It is guided by teachers but substantially student controlled. It incorporates multiple ways to demonstrate learning over time. This encourages pupils to build on their interests. It also provides the basis for making decisions about how best to improve student outcomes, teaching and schools.

In FairTest’s model, states design a “system of systems.” Districts, or consortia of schools or districts, have the flexibility to ensure that the structure and nature of their assessment systems address local needs and challenges. This could range from assessments rooted in inquiry- and project-based learning, with extensive student choice, to more traditional curriculum, instruction and tests.

To fulfill ESSA’s public reporting and accountability requirements, the model system relies primarily on classroom-based evidence. Teachers and students gather examples of learning throughout the school year. Teachers prepare a summative evaluation of each pupil. This includes a determination of the student’s level of proficiency, based on state standards, as required by federal law. The data are aggregated and then broken out by demographic groups. The results can shed light on the success or failure of efforts to close gaps in achievement.

ESSA requires such new state systems to establish “comparability” across schools and districts. A state must determine whether students deemed proficient in one district would receive a similar evaluation in another with a different local assessment system. Typically, this involves using state standards as the basis for re-scoring samples of classroom-based work. This, in turn, provides information needed for public reporting and accountability actions.

FairTest’s model is anchored in a set of assessment principles. It also reflects real-world experience and evidence. For example, New Hampshire is entering the third year of its Performance Assessment for Competency Education (PACE) pilot program. PACE combines limited state testing with teacher-made Common Tasks used across districts to establish comparability. School assessment systems include teachers evaluating each student based on local tasks, plus a complete review of student work over the year. New Hampshire’s Rollinsford Grade School demonstrates how school-based assessing can succeed when the evidence of learning emerges from ongoing work guided by teachers but significantly determined by students.

Other examples of performance assessment presented in the report include the New York Performance Standards Consortium, Learning Record, Work Sampling System, Big Picture Learning, and International Baccalaureate Program. They demonstrate that performance assessments can significantly improve the chances that disadvantaged students will overcome obstacles and reach their potential. They also illustrate ways schools and districts can show comparability.

FairTest’s model is intended to help states design flexible systems that provide accountability while ensuring that accountability structures do not undermine rich teaching and learning. Some other ESSA requirements may create difficulties in implementing quality assessment for learning. Nevertheless, the space for progress is large enough to make the innovation pilot an important step forward, if used well.