Rubrics are explicit grids, schemas, lists, to evaluate and classify products of learning outcomes assessment into different categories that vary across a given continuum.
Rubrics are handy in measuring student learning given some specific learning outcomes.
There are different ways to build a rubric, often the building process takes some thought and time, but the rubric gives a clear picture to the student and instructor about the level of performance expected for a certain task.
Rubrics are criteria that cover the essence of a performance that is judged with them. Rubrics help define expected performance, standards of quality, levels of accomplishment. They are also helpful in diagnostic assessment and providing feedback to students.
Holistic vs. Analytical
Holistic rubrics give a single score or rating for an entire product or performance based on overall impression of a student’s work.
The rater considers all quality judgments in one big component and overall judgment and comes up with one single score.
- Student shows complete understanding of the tasks and concepts
- Clear identification of key concepts and important elements
- Excellent writing style
- Pertinent insight and demonstration of appropriate application of main ideas
- Understanding of most critical concepts
- Shows identification of some key concepts but most of the parts are missing
- Adequate writing style with minor errors, some limited clarity in expressions
- Scarce demonstration of application of main ideas
- Misunderstanding of majority of concepts or no understanding of concepts and processes
- Irrelevant or illegible response that has no relation to the key concepts
- Unsuccessful attempt to communicate
- Lack of demonstration in application of main ideas
Holistic Rubrics Are Suitable for …
- Judging simple products or performances
- Getting a quick snapshot of overall quality or achievement; often used when a large number of students are graded
- Judging the impact of a product or performance more than the specific detailed parts of the performance.
There is no detailed analysis of the strengths and weaknesses of the performance or product, so holistic rubrics are not useful as diagnostics or for giving students detailed feedback on their performance. Holistic rubrics offer little in the way of help to students who would improve their performance.
Analytical rubrics divide a product into essential dimensions (traits), and each dimension is judged separately. A separate score is given for each dimension or trait considered important for the assessed performance.
Scoring of each trait can be done by using a Likert scale (e.g., 1 to 5 where 1 is poor quality, 3 is average, and 5 is excellent quality).
Example—Six-trait Analytical Writing Rubric
Each of the following traits is scored separately on a scale of 1 to 5:
Ideas: the main idea of the assignment, content ideas, main theme, details in the theme
Organization: the internal structure of the writing, the core and central meaning, logical and creative pattern of ideas
Voice: the feeling and beliefs of the writer come through the words
Word choice: rich, precise, scholarly, and eloquent language that enlightens the reader
Sentence fluency: the rhythm and flow of the language, the word patterns, the way the writing plays to the ear
Conventions: the mechanical correctness of the work as an entire piece, along with grammar and usage of words (spelling, capitals, punctuation).
Analytical Rubric Are Suitable for …
- Judging complex performances that involve multiple dimensions (skills that must be assessed). Each step in the rubric can be designed to measure one specific trait.
- Provide more specific information and feedback to students about their strengths and weaknesses.
- Can be used to target instruction to specific areas in need for improvement.
- Analytical rubrics help students come to a better understanding about the nature and quality of work they must perform.
- More time consuming to craft and use in grading
- Lower inter-rater agreement because of the many and detailed traits
- Less desirable in large scale assessment context when many students must be graded and when speed in grading is essential
General vs. Task-Specific Rubrics
General (generic) rubrics can be used across similar performances (e.g., all Blackboard postings across a semester; all group interactions in the semester). Construct one rubric, and use it for similar tasks. Sometimes the tasks are so similar that there is no reason to construct a detailed rubric for each.
- When grading portfolios or any collection of different assignments that must be graded overall, or where developing a rubric for each piece would be time consuming and not important.
- To help students understand the big picture and when students need to be able to apply what they learned in one task to the next task
- When the product is more important for assessment purposes than the process/steps of that performance
- When the students must come up with general ideas and when they must think in order to develop particularized performance
- When there are many different possibilities to solve a problem or there are many ways for a successful performance
Task-specific rubrics can be used only for a particular task or assignment (e.g., rubric for the Final exam,; rubric for a project proposal).
- When it is easier and more consistent to get fast scoring. Easier to train the scorers
- When you want to know if students are able to perform specific methods and procedures
There are times when the best rubric is a combination of generic and task-specific rubric traits. For example: when you want to see whether or not students know how to develop specific parts or steps in a project proposal but, at the same time, you are also interested in the overall quality of writing.
Number of score points for rubrics
- Sufficient range to evaluate the performance
- Enough points to evaluate different levels of quality
- If your rubric needs to fit a particular standard, use the same amount of points.
- Tracking changes over a longer period of time requires more points than a limited or shorter time.
- Have good reasons to pick odd or even number of points: odd numbers have a tendency to gravitate to the mean or a more neutral assessment; but, if your goal is to delimitate the average, then an even number of points is the best choice.
- 1-5 point scale is similar to A-F grading, at times students confuse points with grades
Arter, J. & McTighe, J. (2001). Scoring rubrics in the classroom. Thousand Oaks, CA: Corwin.
Brennan, R. L. (Ed.) (2006). Educational measurement (4th Ed.). NCME. Westport, CT: Praeger.
Earl, L. M. (2003). Assessment as learning. Thousand Oaks, CA: Corwin.
Huba, M. E. (2000). Learner-centered assessment on college campuses. Boston, MA: Allyn & Bacon.