MIT Visualizing Cultures


Asia Rising and Yellow Promise/Yellow Peril Curriculum

RUBRICS: General Principles and Guidelines


1. A rubric:
a. Contains a scale of possible points for scoring work on a continuum. High numbers usually are assigned to the best performances.

b. Provides descriptors for each level of performance assessed.

c. Is either holistic or analytic. If holistic, the rubric has only one general descriptor for the performance as a whole. If analytic, there are multiple rubrics corresponding to each dimension of performance or trait being scored. (e.g. organization, ideas, voice, sentence fluency.)

d. Should be tailored to the assignment.

e. Is available to students from the beginning to the end of the assignment.
2. The rubric must enable teachers and students to effectively discriminate between performances of different quality.
a. The discrimination must be valid: the characteristic differences of each level of performance must be salient, not arbitrary.

b. The discrimination must be reliable: The scores by the same teacher at different times or different teachers at the same time must be consistent within reasonable limits.

c. Rubrics tend to sacrifice validity for reliability. Beware!
3. Scoring should alert students to their real level of performance. It may well happen that no one gets the highest score; it may happen that many students get low scores. The score is not based on who does the best work in the class (norm-referenced); it is based on how well students perform compared to the criteria (criterion-referenced).

4. The rubric will contain rich descriptors of levels of performance if the descriptors are generalizations derived from actual examples of student work.

5. The number of points on the scale is typically between four and six for rubrics used in classroom-based assessment. An even number of points is better than an odd number because it forces more care in judging.

6. The more task-specific the rubric, the more valid the result; the more clear and simple the rubric, the greater the reliability.








Massachusetts Institute of Technology © 2008 Visualizing Cultures