This article is still a work in progress, expected to be finished within the next weeks.
Rubrics have become a staple in education, and for good reason, boasting a multitude of benefits for both summative and formative purposes. For this reason, we want to provide helpful guidelines on when to use which kind of rubric, as well as approaches to constructing well-designed rubrics.
The article will cover four main topics:
The basic architecture of rubrics
Three common misconceptions about rubrics.
Two approaches to rubrics
Components of good criteria and performance level descriptions
The basic architecture of rubrics
Rubrics should always consist of criteria and descriptions of the level of performance across the criteria, how many of these you need depends on the rubric design you choose.
There are two main types of rubrics, holistic and analytic, and each of these can be executed as either a general or task-specific rubric.
You can use the decision tree below to see which is most appropriate for your use, along with the pros and cons of each type.
Common misconceptions about rubrics
Confusing learning outcomes with tasks : Rubrics should not be assignment directions. Instead of being task-focused, they should focus on the proficiencies that we should expect to see students display after having finished the task. This way we can also help students focus on the learning outcomes, rather than only completion and grading.
Confusing rubrics with requirements of quantities : Instructors must be careful to avoid the rubric becoming a sort of checklist. In practice, it’s highlighting that students should display the ability to conduct thorough research, considering multiple criteria when selecting information sources, instead of requiring students to use five sources to score high.
Confusing rubrics with evaluative rating scales : Rubrics help structure observation and provide students with a description of where they currently stand. The rating scales should therefore be descriptive instead of numerical so that it facilitates the bridge between what students display and the judgement of that, but also helps students easily envision the next steps in their learning process, rather than focusing on their numerical grade score.
Two approaches to constructing rubrics
To avoid these common mistakes with rubrics we suggest two approaches: adopt and adapt and bottom-up.
Adopt and Adapt
The approach sometimes called Adopt and Adapt, literally means adopt a rubric you found online, or that was shared by your colleagues, and adapt it to your teaching context .
If you find a rubric, say for a research paper, then there are a few steps you can take to best adapt it to your and your student’s needs.
Step 1: Consider the assessment task and the context in which the rubric will be applied
Step 2: Evaluate whether the criteria can be kept as they are, for example, a criterion on “Explanation and summary of the issue” may still be relevant for your learning outcomes on your assessment too, so you may decide to keep it.
Step 3: Consider whether the description level performances of the criteria in the rubric you found suit your particular learning context. In this case the description covers most relevant parts, but imagine you would want your assignment also is more opinionated, so you encourage a statement of position. Well then you can go ahead and add that to the description, just like below.
In general, the process of adapting rubrics should be iterative over time.
The bottom-up approach is an inductive process, which begins by sampling student work and using them to create a framework for assessment .
Step 1: Start by collecting up to a dozen of student work examples. These samples should all relate to the same overarching topic but should be different examples because the rubric should reflect the learning outcomes rather than the specific task.
Step 2: Sort the work into three different levels: high, medium and low level.
Step 3: Write descriptions of why each piece was sorted in the way it was. What made this paper this level, maybe the synthesis of information was thorough and creative.
Step 4: Compare and contrast those descriptions of the work, if some descriptions come up often, e.g. “includes research to support claims well” then the ability to integrate information may emerge as a dimension.
Step 5: For each of the criteria identified in the previous steps, try to write descriptions of performance along the dimensions (Beginning, Emerging…) for the appropriate amount of levels. There is no right number of levels but keep in mind that you should be able to clearly distinguish between each level, and they should be useful in guiding observations.
Criteria and performance level descriptions
It is important that your criteria meet these requirements :
Appropriate: Each criterion should represent an aspect of the goal or objective that we intend for students to learn, for example, if it's teamwork, their contributions to the team's work may be a relevant criterion.
Definable: It is important that each criterion consists of clear and agreed-upon meaning, that both teachers and students understand.
Observable: Criteria are used to guide judgement on learning, so to be useful and valid they should be something that teachers or anyone reviewing the work can observe. This could be in the form of behaviour in teamwork, for example, inclusivity.
Distinct: It is important that each criterion is unique, if we cannot distinguish between them they lose part of their value. It also becomes more difficult for students to assess and differentiate their learning progress if criteria overlap.
Complete: Each criterion on its own may not represent the learning outcomes intended, but when taken all together the criteria should represent all the areas of learning and skills that we aim to foster in students.
Support descriptions along continuums: It should be possible to describe the criteria over a range of levels, if this is not feasible then the criteria may not be ideal for use.
Figure 2: Example of criteria for a collaboration rubric.
Your descriptions of the level of performance should meet these requirements :
Descriptive: Performance is described in terms of what is observed in the work. Avoiding evaluative language is key.
Clear: Mutual understanding between student and teacher.
Cover the whole range of performance: Performance is described from one extreme of the continuum to another for each criterion.
Distinguish among levels: It is important that the descriptions betweens levels are different enough so that there isn’t space for ambiguity on both the student and teacher's side when matching examples of work to performance descriptors.
Matching the target performance to the appropriate level: Make sure that the description of performance level is an appropriate level that one can expect from students, given the goal of the lesson, assessment or curriculum.
Feature parallel descriptions from level to level: Across the rubric for each criterion, there should be descriptions of performance, this should represent a continuum ranging from low to high.
Two common ways to start writing out the descriptions across the levels are:
Start with the level you expect most students to meet, this might for example be the level "Proficient", then go from here, backing down towards the lowest, or vice versa.
Start with the highest level, in this example "exemplary" and then downwards from there.
Figure 3: An example of descriptions of performance, for visualization purposes only two descriptions are provided.
 Brookhart, S. M. (2013). How to create and use rubrics for formative assessment and grading. Ascd.
 Ragupathi, K., & Lee, A. (2020). Beyond fairness and consistency in grading: The Role of Rubrics in Higher Education. In Diversity and inclusion in global higher education (pp. 73-95). Palgrave Macmillan, Singapore.