Please review these 5 suggestions for crafting a strong Rating Rubric. Following these suggestions will encourage students to provide constructively critical reviews and increase the likelihood that your assignment will produce strong Reliability Ratings for each Reviewing Dimension.
- Use concrete terms (e.g. The paper provides necessary definitions clearly and consistently) rather than generic descriptors (e.g., Good, Poor) in your Rating Rubrics.
- Create multiple Rating Rubrics if you wish to assess students on multiple aspects of their work. It is better to have two different Rating Rubrics and have students assess both features of the assignment, because it is confusing for students to deal with cases that are high on one element but low on the other if they are asked to evaluate these on the same rubric.
- Keep your explanations for the Ratings concise. If you enter long explanations for all 7 levels, this will be hard to read. Some instructors use sentence-long anchors for levels 1,3,5,7 and put number anchors only for 2,4,6.
- Remember, if you leave a middling entry blank, the scale will omit that level entirely. For example, if you only put values in 1,3,5,7 with the rest blank, then students will have a 4-point scale.
- Use a spread of ratings across the scale of 1-7. For example, if you wish to use a 4-point scale, do not just pick 1,2,3,7 because that will give very strange average scores. It is better to use 1,3,5,7.