Writing to the Rubric: Lingering Effects of Traditional Standardized Testing on Direct Writing Assessment.

Abstract

Performance assessment offers a way out of the factory and into the classroom, the author points out. But to get there, we need to reconsider our mania for making test-based comparisons that demand standardization. AS SHE placed test materials, dictionaries, and thesauruses strategically on empty desks in preparation for administering Pennsylvania's state-mandated direct writing assessment, a middle school teacher awaiting her students said: See this box of laminated rubrics? That's the rubric the state uses to score this test. I also score the writings of my students in this class with it. Sometimes I have students use it to score each other's papers. They have to justify the scores they give, so they are very familiar with this rubric. They know what they have to do today when they take the test. But the test is not really a fair representation of how well they can write. They could do better if they had a choice of topics and could do the sorts of things we usually do. For instance, with these papers over here, they picked a topic one day in advance and could think about it, look at some reference materials, take some time planning and drafting, and come up with a polished product. I was very pleased with these papers. I'm frustrated with the test. We've done all of this preparation, and I've organized to the max so they can concentrate on their writing during the test time, but my students will still score at only about the state average. We're a rural district, and we don't have all the curricular options you find some places. I have to teach literature and grammar as well as writing, and I throw in spelling they're not always going to have a spell-checker handy. But my students will be compared to some suburban students who are in a writing-only curriculum. When the scores are printed in the newspaper, people will think we're not doing a good job of teaching here. It just doesn't seem fair.1 The direct assessment of student achievement in writing judging writing skill on the basis of actual student writing instead of multiple-choice test items was supposed to fit better with daily classroom activities and objectives and was supposed to provide a better basis for making valid inferences about student learning. So what has gone wrong when a state-mandated performance assessment is threatening curriculum, creating anxiety, and raising equity concerns? As part of the larger argument that traditional psychometric concepts and practices undermine new ideas and techniques in assessment, I will argue that rubrics have the power to undermine assessment. Scoring rubrics are pivotal in operationalizing large-scale and standards-based performance assessments in writing. Rubrics promote reliability in performance assessments by standardizing scoring, but they also standardize writing. The standardization of a skill that is fundamentally self-expressive and individualistic obstructs its assessment. And rubrics standardize the teaching of writing, which jeopardizes the learning and understanding of writing. Operationalizing Standards In Rubrics Of the performance assessments in state testing programs, direct writing assessment is the most prevalent and longest-running.2 Currently, 38 states assess students' writing skills through direct writing assessments, and all 38 use a rubric to score students' performances.3 Rubrics. Frequently referred to as scoring guides, assessment rubrics are rules by which the quality of answers is determined. Each rubric is a description of student performance that clearly articulates the requirements for each of the score points.4 Rubrics are translations of visions of desirable performance into specifications of exactly what is desirable. The use of rubrics to score performance assessments predates the current standards movement, but rubrics are crucial in current implementations of academic standards. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 85 citations exist upstream.