Summary findings from "Using Rubrics to Respond to Student Writing."
Why use a rubric?
As labor saving devices, rubrics are of questionable value (at least
if you spend as much time tinkering with them as I do). However, most
of us agreed that their real value is as a tool for more clearly communicating
expectations and feedback to students, and in assisting us to grade
more consistently (especially at 11 p.m. on Sunday night).
Types of rubrics:
- Analytic rubrics allow you to evaluate the project's performance
in individual categories such as argument, organization, and so on.
Descriptions of criteria (or examples) are given for a range of performance
levels from superior to unacceptable. Typically, each category is
given a weighted score. When done well, this type of rubric provides
fairly concrete feedback to the student, and it encourages consistent
grading. On the other hand, weighted scores sometimes introduce the
risk that a student can be disproportionately penalized or rewarded
for performance in a given category. As Robert Miller said, a passing
paper can be beautifully written...and completely wrong.
- Holistic rubrics generally have three or four score levels,
each of which provides a description of the writing components that
contributed to the score. For instance, a level 3 (with 3 being the
highest score) might include the following description:
Meets All Expectations- Introduces the topic by engaging the reader in an interesting anecdote.
- Uses good transitions to move the reader effectively from personal recollection to broader ethical point.
- Documents sources completely and correctly using MLA or APA documentation standards.
- Concludes the paper with a provocative rhetorical question.
Level 2 would presumably give descriptions of each component that were merely satisfactory. While holistic rubrics may be fast to use, they unfortunately do not allow for a particularly nuanced response to student writing.
- Hybrid rubrics generally borrow the structure of analytic rubrics for the purpose of description, but they may not score each category. This eliminates the problems associated with weighting scores, but students who wonder why they received a particular grade may not find them as useful. In a sense, these are feedback mechanisms rather than scoring rubrics.
Considerations as you develop a useful rubric:
- One thing we all agreed on was that rubrics should be designed in tandem with the assignment. As the assignment describes the criteria for a successful project, the rubric should reflect (and perhaps even clarify) those same values.
- Likewise, some suggested that the rubrics should use the same language as ("link to") the overall course objectives (in addition to the assignment goals).
- Be sure to "unpack the language" by avoiding ambiguous terms like fluency, voice, sufficient, acceptable, and so on, in both the rubric AND the assignment.
- A few people mentioned that they develop their rubrics collaboratively with their students. Robert Miller, for instance, even allows his students (with his guidance and, if necessary, veto power) to assign the values for each category of the rubric. Certainly, that strategy implies that students better understand the language and values of the rubric and assignment.
- What would an exemplary example of this assignment look like? A satisfactory example? Unsatisfactory? Roy mentioned that he prefers to label his lowest category "Not Yet Acceptable."
- Consider how your rubric/assignment can challenge students. Where do you want them to develop or improve? Elissa suggested it might be appropriate to develop individual rubrics for each student depending on their level or goal.
- Consider asking your students to rate their own paper (or sample papers) using the rubrics.
