Skip the Scantron: Why Evidence-Centered Design is a Better Testing Option
In an era of increased standardized testing and a focus on standards and competencies, teachers must use the best methods for measuring student learning. The rote memorization and multiple-choice tests of years past are increasingly incapable of assessing the complex tasks modern students perform.
These old testing strategies are too narrow for the high-level benchmarks we need to meet. The new assessment challenge lies in examining students’ reasoning, understanding and ability to apply “authentic knowledge.” With that in mind, educators use essay exams, exit tickets and a wide variety of assessments to help them get snapshots of student learning. Another dynamic assessment technique is evidence-centered design, or ECD.
How evidence-based testing works
Developed in the late 1990s by Robert Mislevy, Linda Steinberg and Russell Almond, ECD identifies variables in student learning and understands that competency flows from a balance of these variables.
Consider the ECD approach to a math test. It doesn’t work like the old assessment designs that might give students a multiple-choice or short-answer question, seeking a final, calculated result. Hurried or careless students might know how to answer the question correctly but get it wrong on the test because of a transcription error or miscalculation. Indeed, show-your-work problems could be counted wrong if the student failed to provide the appropriate short answer or formula, regardless of the calculated result. The old measures don’t necessarily illustrate student reasoning or authentic knowledge.
ECD gives instructors opportunities to measure both process and result, enabling the aforementioned hypothetical students to receive some credit on a question even if their final answer is incorrect. If the instructor identifies process and product as two variables, the assessment expands to account for student thinking in addition to computation. This also allows for multiple routes to an answer, and the option of rewarding variety and complexity if the teacher so chooses.
Such assessment is the norm for English instructors, who often joke about wishing they gave multiple-choice tests. Grading an essay means measuring the balance of student proficiency in a variety of variables, including grammar, thesis, audience or structure.
Common traits of ECD
Regardless of content discipline, there are clear commonalities in developing ECD. Such assessments are not narrow or rigid and must be thoughtfully and carefully designed and applied with a deep understanding of the various means of measuring student competency. Because students are unaccustomed to such measures, clear rubric design and communicating instructor expectations to students are key to avoiding issues of subjectivity.
While evidence-centered design requires thoughtful deliberation, it yields significant results. In her recent piece “Evidence Centered Assessment,” Kimberly Morrow-Leong highlights the distinct advantages to such an approach in mathematics, saying that it restructures her thinking on assessment from a “what’s wrong” to a “what’s right” approach, giving her an opportunity to easily identify gaps in student knowledge and how to address them with future instruction.
In this way, evidence-centered design is a much more significant tool than we ever imagined, allowing instructors to not only assess student learning, but to specifically and continually craft curriculum and course design that addresses learners at the classroom and individual level.
Such nuanced understanding and instruction can acknowledge to students what teachers have always known: They are far too complex to be measured by picking A, B, C or all of the above.
Monica Fuglei is a graduate of the University of Nebraska in Omaha and a current faculty member of Arapahoe Community College in Colorado, where she teaches composition and creative writing.Learn More: Click to view related resources.
- Valerie Shute, Yoon Jeon Kim, Rim Razzouk, "ECD for Dummies"
- Kimberly Morrow-Leong , "Evidence-Centered Assessment," National Council of Teachers of Mathematics