“Rubrics are a popular approach for focusing learning and for assessing and reporting student achievement,” writes Rick Wormeli is his recent book Fair Isn’t Always Equal: Assessing and Grading in the Differentiated Classroom. “Designing rubrics may be more complex than teachers realize,” Rick continues, “however, but we get better at it with each one we do.” And to help with that practice, he outlines seven steps to designing an effective, useful rubric.
How to Design a Rubric
1. Identify the essential and enduring content and skills you will expect students to demonstrate. Be specific.
2. Identify what qualifies as acceptable evidence that students have mastered content and skills. This will usually be your summative assessments and from these, you can create your pre-assessments.
3. Write a descriptor for the highest performance possible. This usually begins with the standard you’re trying to address. Be very specific, and be willing to adjust this descriptor as you generate the other levels of performance and as you teach the same unit over multiple years. Remember, there is no such thing as the perfect rubric. We will more than likely adjust rubrics every year they’re used.
4. At this point, you’ll have to make a decision: holistic or analytic? If you want to assess content and skills within the larger topic being addressed, go with analytic rubrics. They break tasks and concepts down for students so that they are assessed in each area. Analytical rubrics also require you to consider the relative weights (influences) of different elements. For example, in an essay, if “Quality of the Ideas” is more important than “Correct Spelling,” then it gets more influence in the final score. If you want to keep everything as a whole, go with holistic rubrics. Holistic rubrics take less time to use while grading, but they don’t provide as much specific feedback to students. In some cases, though, the difference in feedback is minor, and the work inherent with an analytical rubric doesn’t warrant the extra time it takes to design and use, especially at the secondary level where teachers can serve more than 200 students.
Another way of looking at the difference is this: The more analytic and detailed the rubric, the more subjective the scores can be.
The more gradations and shades of gray in a rubric, the more the score is up to the discretion of the teacher and is likely to differ from teacher to teacher, and even from day to day. The more holistic the rubric, the fewer the gradations and shades of gray and thereby, the more objective and reliable the scores can be. Of course, the more detailed the rubric, the more specific feedback we get for both teacher and student. It’s very rare to generate a rubric that is highly detailed and analytical while remaining objective and reliable teacher to teacher and over time.
Here are two examples: In a holistic rubric, we might ask students to write an expository paragraph, and the descriptor for the highest score lists all the required elements and attributes. With the same task in an analytical rubric, however, we create separate rubrics (levels of accomplishment with descriptors) within the larger one for each subset of skills, all outlined in one chart. In this case, the rubric might address: Content, Punctuation and Usage, Supportive Details, Organization, Accuracy, and Use of Relevant Information.
In a chemistry class’s holistic rubric, we might ask students to create a drawing and explanation of atoms, and the descriptor for the highest score lists all the features we want them to identify accurately. With the same task using an analytical rubric, however, we create separate rubrics for each subset of features—Anatomical Features: protons, neutrons, electrons and their ceaseless motion, ions, valence; Periodic Chart Identifiers: atomic number, mass number, period; Relationships and Bonds with Other Atoms: isotopes, molecules, shielding, metal/non-metal/metalloid families, bonds (covalent, ionic, and metallic).
Remember how powerful this becomes when students help design the rubric themselves. After working with a few rubrics that you design, make sure to give students the opportunity to design one. Determining what’s important in the lesson moves that knowledge to the front of students’ minds, where they can access it while they’re working. This happens when they have a chance to create the criteria with which their performances will be assessed.
5. Determine your label for each level of the rubric. Consider using three, four, or six levels instead of five for two reasons: 1) They are flexible and easily allow for gradations within each one, and 2) a five-level tiering quickly equates in most students’ and parents’ minds to letter grades (A, B, C, D, F) and such assumptions come with associative interpretations—the third level down is average or poor, depending on the community, for instance. The following list shows collections of successful rubric descriptor labels. Though most are written in groups of five, which I advise teachers not to use, they are provided in such groupings because that is what educators most commonly find on their district assessments. Look at the list’s entries as a sample reservoir of word choices.
- Proficient, capable, adequate, limited, poor
- Sophisticated, mature, good, adequate, naïve
- Exceptional, strong, capable, developing, beginning, emergent
- Exceeds standard, meets standard, making progress, getting started, no attempt
- Exemplary, competent, satisfactory, inadequate, unable to begin effectively, no attempt
Descriptor terms need to be parallel; it’s important to keep the part of speech consistent. Use all adjectives or all adverbs, for example, not a mixture of parts of speech. Notice how this sequence on a rubric could be awkward for assessment and confusing to students:
- Top, adequately, average, poorly, zero
6. Write your descriptors for each level, keeping in mind what you’ll accept as evidence of mastery. Once again, be specific, but understand that there is no perfect rubric. Alternative: Focus on the highest performance descriptor, writing it out in detail, and then indicate relative degrees of accomplishment for each of the other levels. For example, scoring 3.5 on a 5.0 rubric would indicate adequate understanding but with significant errors in some places. The places of confusion would be circled for the student in the main descriptor for the 5.0 level.
In my own teaching experience, this alternative has great merit. When students are given full descriptions for each level of a rubric, many of them steer themselves toward the second or third level’s requirements. They reason that there’s no need to be “exemplary”— the top level—when they’d be happy with the label “good” or “satisfactory.” These students either don’t believe themselves capable of achieving the top score’s criteria, or they see the requirements as too much work when compared with the lower level’s requirements. To lessen the workload, they are willing to settle for the lower-level score.
Don’t let them do this; don’t let them lose sight of full mastery. When all that is provided to students is the detailed description of full mastery, they focus on those requirements—it’s the only vision they have. All of their efforts rally around those criteria and, as a result, they achieve more of it.
7. “Test drive” the rubric with real student products. See whether it accounts for the variable responses students make, ensuring those who demonstrate mastery get high scores and those who don’t demonstrate mastery earn lower scores. Ask yourself: “Does this rubric provide enough feedback to students to help them grow? Does it assess what I want it to assess? Does it help me make instructional decisions regarding students’ learning?” If it doesn’t do one or more of these things, the rubric may need to be reworked.
Add comment July 20th, 2010