Rubrics for graded assessment

Ross Woods, 2018

The purpose of this paper is to give guidelines for graded assessment rubrics in postsecondary education.

In graded assessment, student work is given an assessment result such as A, B, C, D, and E. Rubrics can be used to differentiate between levels of performance required for different grades. A rubric is a set of specific criteria so that student work can be assessed more objectively, consistently, and easily. Rubrics for graded assessment give separate criteria for each grade and are often presented in table formats.

 

What grades?

For the purposes of this paper, grades are as follows:

AOutstanding: Good performance at the next highest qualification.
BVery good: Passing at one qualification level higher than that for which it is offered. It also demonstrates the ability to continue to the next highest qualification.
CSatisfactory, adequate performance
C-Minimum passing requirement
DDoes not meet minimum passing requirement
EVery poor

 

In many assessment systems, these may be modified with plus or minus, resulting in the grades A, A-, B+, B, B-, C+, C, C-. (Some systems have A as the highest possible grade, while others include a grade of A+.)

Two particular grades have critical roles:

 

Good practice: Competency based assessment and grading

  1. Graded assessment is not equally suitable for all units. Units are best assessed as Competent/ Not yet competent if they refer to competencies that are either you can do it or you can’t. Either way, all students in the institution should be assessed in the same way for a unit. If it is graded, then all should be graded. If it is pass/fail, then all should be pass/fail.
  2. Students must achieve the minimum requirement for all unit requirements to gain a grade of C- or above. That is, high achievement in some requirements cannot compensate for inadequate performance in others.
  3. Graded assessments are subject to the same validation (moderation) and appeal procedures as non-graded assessment.
  4. Like any assessment, the reasons for the grade should be documented.
  5. It is not good practice to cluster units, assess them together, and give one assessment result for the whole cluster. Each unit should be given a separate grade because levels of performance might differ between units.
  6. Grading in competency-based assessment has no business with some assessment practices:
    1. taking marks off for errors and giving extra marks for very good work.
    2. percentages, which are arbitrary conventions, although they may be mapped to assessment grades.
  7. Students are entitled to the disclosure of assessment standards. If students are given rubrics, they can plan their work to achieve at higher levels.

 

Methods of writing rubrics

1. Collate existing assessment comments. If assessors have given reasons for their assessment decisions as they go, it is a matter of collating those comments into one document and editing it into a consistent document. This uses criteria that have actually been actually applied. If assessors use a Learning Management System, the software can assemble those assessment comments. If assessors keep individual manual records, the act of collation must also be manual.

2. Build a rubric as you go. In this approach, assessors keep notes of factors affecting grades when they make assessments and validate (or moderate) assessments. It is most useful if assessors record assessment decisions of borderline cases, and list the reasons for giving the assessment result that they did. Rubric writers can write as they go in several ways:
• Put them into a structured document as they go.
• Annually collate them and issue (or update) a rubric document.
• Collaborate with other assessors to collate notes and annually issue (or update) a rubric document.

3. Interview assessors. Get their views on assessment requirements and collate them into a consistent rubric document.

4. Research existing rubrics. Don’t re-invent the wheel. Using existing rubrics has the advantage of learning from the lessons of others. However, they aren’t always ideal. Copyright is the most obvious barrier. Existing rubrics might not suit the particular needs of your institution or context. In some cases, the criteria in existing rubrics might anticipate assessment needs but might never have actually been actually applied. In other words, they might be good theory but not reflect actual practice.

 

Generic examples of rubrics

The grade of E. Both examples below omit the grade of E. In these examples, the grade of E indicates that the task was either not done at all or done so poorly that it might as well not have been done.

Sudden death. Some assessments have a sudden death factor, aso known as a fatal flaw. If students get something essential wrong, they must be be given a grade of D nor E, no matter how well they do everything else in the assessment. This is different from a minor error made by a competent student who still has room for improvement. Here are a few examples:

  1. If you break the law only once in a driving test, you've failed the test.
  2. If a student's essay is built on a major contradiction or fallacy, you can't give it a passing grade. In this case, a major contradiction or fallacy is different from a minor error that does not affect the conclusion.
  3. If a student acts dangerously in a real workplace during an assessment, he/she has already failed the assessment. You may even need to stop it. In this last case, acts dangerously is different from having a minor risk that does not seriously compromise safety.
Example 1: Practicum or project
A B C D
Demonstrates initiative and creativity Occasional initiative and creativity Only follows instructions Does not always follow instructions
Consistent strong leadership Show strong leadership Occasional leadership Does not show leadership, or bossy/dictatorial
Sets an example in professional skills and behavior, able to teach others Can do the job with negible, inconsequential errors Can do the job, occasionally makes small errors Makes frequent small errors
Made a critical (fatal) error
Permitted only to work under supervision
Gets all work done on time, even under pressure Gets all work done on time Gets all work done on time Often does not get tasks done on time
Can work independently Can work independently Can work independently Unable to work independently; requires supervisor support and correction

 

Example 2: Essay requiring critical thinking and argument
A B C D
Creative topic that does not replicate existing papers Well-chosen topic Routine topic chosen according to instructions Follows routine topic generated by instructor; lacks innovation and creativity
Conclusion is firmly proven Adequate argumentation and data to support conclusion Plausible conclusion. Any errors in logic do not invalidate conclusions Conclusion not valid due to dependence on a logical fallacy
Publishable quality of language and presentation Minor errors in layout, grammar, and/or language style Readable, but has obvious small errors in layout, grammar and/or language style Multiple errors in layout, grammar and language style that detract from readability
Original contribution to knowledge in the field or to professional practice Engage with significant topic Engages topic Tends to "miss the point"
Argues convincingly against opposing views Engages opposing views Notes opposing views Ignores opposing views
Uncovers hidden implications Explores implications Notes some obvious implications Ignores implications
Proposes innovative or creative solution Proposes workable solution Offers obvious solution Unclear conclusion
Very good structure Good structure Structure is weak but still understandable Jumbled structure
Good range of supporting information or data Good range of supporting information or data Adequate information or data Inadequate information or data (e.g. only one or two sources)
Quotes and referencing all correct Quotes and referencing all correct Minor errors in quotes and referencing Has fatal flaws: plagiarism, misquote, missing references