Rubrics

Concrete Grading Criteria

Concrete grading criteria are criteria that can be viewed and assessed with less interpretation and subjectivity. Examples include:

• Content knowledge or declarative knowledge (about a topic or learning objective)
• Procedural knowledge (knowledge about how to do a task or action)
• Conditional knowledge (knowledge about why or when to do an action)
• Art composition
• Argument with justification or defense
• Accuracy or correctness
• Information literacy (supporting ideas with research and creating new information from research)
• Writing mechanics (spelling, grammar, punctuation, capitalization)

For example, you might develop a rubric or checklist for weekly math assignments that includes grading criteria for procedural knowledge (showing work), conditional knowledge (explaining why they used a formula or operation), and accuracy (correctness of answer).

Abstract Grading Criteria 

Abstract grading criteria are grading criteria that are interpreted and are considered more subjective than concrete grading criteria. Examples include:

• Critical thinking
• Creativity
• Problem-solving skills
• Decision-making or reasoning skills
• Communication or expression of ideas
• Development of new ideas
• Organization or cohesion of writing

For example, you might develop a rubric for a piece of art that includes concrete grading criteria for procedural knowledge (demonstration of specific technique), composition of the piece, as well as abstract grading criteria for creativity and decision-making skills.

It is important to note that abstract grading criteria can be difficult for students to know what the expectations are and how to demonstrate those expectations. Abstract grading criteria can be described in a rubric to help students understand the expectations.

Rubric performance levels are usually labeled with a column heading that can be a numeric point value, percentage, letter grade, or heading title. For example:

• 100% – A level of performance could use any of the following terms as a heading: Exemplary, outstanding, distinguished, exceptional, excellent, expert, etc.
• 80% – B level of performance could use any of the following terms as a heading: Proficient, above average, accomplished, etc.
• 70% – C level of performance could use any of the following terms as a heading: Satisfactory, competent, average, acceptable, etc.
• 60% – D level of performance could use any of the following terms as a heading: Developing, emerging, approaching, novice, etc.
• 50% – F level of performance could use any of the following terms as a heading: Beginning, rudimentary, needs revision, no evidence, etc.

The above terms can be used as headings for your rubric columns or as adjectives to describe grading criteria at that performance level. It is recommended to utilize the same column headings for all the rubrics developed for a specific course. For example, if you select “Outstanding” for an A level of performance column heading then you should utilize the same column heading for the A level of performance in all your rubrics.

Rubrics contain descriptions of grading criteria. These descriptions should be aligned to the learning objectives being assessed and will support students’ understanding of the assessment expectations. For example, you have the learning objective: Synthesize information and ideas from multiple texts and sources. You label the grading criteria as “Information Literacy” and you describe the grading criteria in an analytic rubric at five performance levels as follows:

• 100% – A level – Outstanding synthesis of information and ideas from multiple credible sources with exceptional cohesion of information presented.
• 80% – B level – Concise synthesis of information and ideas from multiple credible sources with cohesion of information presented.
• 70% – C level – Adequate synthesis of information and ideas from multiple credible courses.
• 60% – D level – Attempted synthesis of information and ideas and/or missing multiple or credible sources.
• 50% – F level – Submission did not demonstrate synthesis of information or ideas, missing multiple and/or credible sources, please revise and resubmit.

See the HOW section of this teaching guide to learn tips for writing criteria descriptions.

Single-point rubrics will have a total number of points or a percentage for the assessment. And each grading criteria in a single-point rubric will have a point or percentage value. Typically, the “meets the criterion” column will be awarded the total points or an A or B value. For example, the assessment is worth 25 points and contains three criteria.

The total points need to be distributed to each of the criteria (criterion I is worth 5 points, criterion II is worth 10 points, and criterion III is worth 10 points). Students who meet all three criteria will be awarded 25 points.

• Small class sizes (under 25 students)
• Involves less time to develop
• Requires more time to grade and score because students need more personalized feedback to understand their performance and score
• Supports conversations about performance

• Can be used for formative and summative assessments
• Appropriate for on-campus or hybrid course modalities
• If using video or audio feedback, it can be adapted for online course modalities
• Best suited for a single user (one instructor)

Analytic rubrics will have a total number of points or a percentage for the assessment. And each grading criteria will have a point or percentage value. For example, the assessment is worth 25 points and contains three criteria. The total points need to be distributed to each of the criteria (criterion I is worth 5 points, criterion II is worth 10 points, and criterion III is worth 10 points). Next, the grading criteria points are broken down further by performance level in an analytic rubric.

• Criterion I is worth 5 points – the highest level is worth 5 points (100%), the next level is worth 4 points (80%), and the last level is worth 3 points (60%).
• Criterion II is worth 10 points – the highest level is worth 10 points (100%), the next level is worth 8 points (80%), and the last level is worth 6 points (60%).
• Criterion III is worth 10 points – the highest level is worth 10 points, the next level is worth 8 points (80%), and the last level is worth 6 points (60%).

• All class sizes
• Involves more time to develop
• Requires less time to grade and score (if the scorer is familiar with the rubric)
• Can be used for formative and summative assessments

• Provides more descriptive feedback in a formative assessment to help students improve performance
• Appropriate for any course modality
• Should be used in online asynchronous course modalities to support student understanding of expectations
• Utilized by multiple instructors and/or TAs

In a holistic rubric, the grading criteria are not broken down and the weighting occurs in the performance levels. For example, the assessment is worth 25 points and contains five levels of performance (the highest level is worth 25 points (100%), the next level is worth 20 points (80%), the third level is worth 15 points (60%), and the fourth level is worth 10 points (40%), and the last level is worth 9 or less points (0 to 39%).

• All class sizes
• Involves less time to develop
• Requires more time to grade and score because students need more personalized feedback to understand their performance and score
• Can be used for formative and summative assessments

• Best suited for summative assessments to measure overall competence or quality of students’ work.
• Appropriate for any course modality
• Best suited for a single user (one instructor)

Checklists will have a total number of points or a percentage for the assessment. And each grading criteria in a checklist will have a point or percentage value.

For example, the assessment is worth 25 points and contains three criteria. The total points need to be distributed to each of the criteria (criterion I is worth 5 points, criterion II is worth 10 points, and criterion III is worth 10 points).

• All class sizes
• Involves less time to develop and grade
• Provides a breakdown of grading criteria
• Used for “Yes or No” or “Correct or Incorrect” performance levels

• Best suited for criteria where there is a correct answer or evidence of correctness or completion: math, engineering, programming, etc.
• Can be used for formative and summative assessments
• Appropriate for any course modality
• Utilized by multiple instructors and/or TAs

Construct underrepresentation refers to when an assessment is too narrow and doesn’t include elements of the construct (course or learning objective). The data collected will not have face, content, or construct validity because the assessment omitted aspects (e.g., the assessment doesn’t capture key aspects of the learning objective it was designed to measure). Content validity refers to how well an assessment measures all facets of an item, and how well it represents or gauges the entire domain (e.g., how well the assessment measures the entirety of the learning objectives). Construct validity refers to how well the assessment collects evidence to support interpretations, appropriateness of inferences, and what the data reflects (e.g., does the data collected allow you to make sound decisions about current instruction or continuous quality improvements).

For example, the evaluation of a piece of art might exclude the composition of the artwork or the grading of an oral presentation might miss the communication of the content (Lin, 2020). The rubric design process helps to ensure that no elements are missing, and all aspects of the construct are being evaluated to improve content and construct validity.

Construct-irrelevant variance refers to when an assessment contains excess or uncontrollable variables that distort the data collected (e.g., the assessment contains grading criteria that are not aligned to the task or learning objectives, or assesses skills and knowledge not taught in the course).

For example, an assessment for an oral presentation has grading criteria for costumes or props. This grading criterion isn’t aligned to the assessment and might cause an assessment bias, a grading criterion that unfairly penalizes students because of personal characteristics (Lin, 2020). In the case of the costume or props criteria, more affluent students could afford better costumes or props and may receive a better grade. This bias would cause an unfairness in grading, and data collected wouldn’t have face, content, or construct validity.

It is essential to review your rubrics to ensure that your grading will be focused on the construct (learning objectives) and that it isn’t missing any elements of the construct or adding any excessive or uncontrollable variables that might distort data or cause an assessment bias.

Internal consistency reliability refers to the interrelatedness of the assessment items and the accuracy of what is measured (e.g., assessments that are directly aligned to the learning objectives would have questions that measure the same construct). Rubric development can enhance the internal consistency reliability of an assessment through the analysis and alignment of learning objectives.

Rater reliability can be described in two sub-categories: intra-rater reliability and inter-rater reliability. Intra-rater reliability refers to how an instructor might grade and score differently based on external circumstances (e.g., one day the instructor is healthy and feeling good and the next day the instructor has a migraine while grading).

Inter-rater reliability refers to how two different instructors might grade and score differently based on what they value (e.g., one instructor might score the organization and technical language in a paper with more weight than another instructor who scores formatting and mechanics with more weight).

Rubric utilization can provide consistent grading criteria that can be repeated under different conditions improving intra-rater reliability (sick instructor) and inter-rater reliability (multiple instructors or TAs). It is important to note that there can still be discrepancies and inconsistencies among multiple instructors or TAs while utilizing a rubric. Please review the HOW section of this teaching guide to learn ways to reduce grading and scoring discrepancies and inconsistencies in order to improve inter-rater reliability.

The following steps will support you as you design a rubric:

1. Title the rubric using the title of the assessment you want to grade and score.
2. Identify the grading criteria that you want to measure. Remember your grading criteria should be directly aligned to the course and learning objectives.
3. Determine how the grading criteria should be assessed: holistically, separately, with a yes/no, etc.
4. Determine how many levels of performance you will measure students’ learning and performance for the identified criteria and what type of feedback you want to provide students.
   1. Single-point rubric – has one column describing a passing performance (typically an A value) and rows for each grading criterion.
   2. Analytic rubric – has between three to five columns to describe performance levels and rows for each grading criterion separately.
   3. Holistic rubric – has between three to five columns to describe performance levels but with only one row as the criteria are described together.
   4. Checklist – has two columns (for yes and no) and rows for each grading criterion.

1. Describe the grading criteria (please see Writing Criteria Descriptions below for more information).
2. Assign points or percentages for each grading criterion (single-point rubric, analytic rubric, or checklist).
3. Describe levels of performance for each criterion and assign points or percentages for each level of performance (analytic rubric or holistic rubric).
4. Review your rubric for mutually exclusive language for levels of performance and student-centered language to ensure student understanding of expectations.
5. Utilize the rubric tool in Blackboard to build a clickable rubric for grading and a viewable rubric for students.
6. Implement the rubric (without making changes) for the entire term. Reflect on the use of the rubric and identify areas of improvement to make adjustments to criteria, descriptions, or weight for the next term.

For more information on building rubrics in your course site visit the Blackboard Grading and Assessments page on the CATE website to view the Getting Started with Rubrics section.

You can co-construct rubrics with students by first sharing a work sample with them. This work sample could be an exemplar (exemplary work sample) or could be an average work sample (B performance level).

The following steps will support you as you co-construct an analytic rubric with your students:

1. Share the course and learning objectives that will be measured by the rubric with students.
2. Share the exemplar (exemplary work sample) or average work sample with students.
3. Break students into groups either synchronously or asynchronously (using a collaborative tool like Jamboard, Google slides, Padlet, Trello, etc.) and ask them to identify what the potential grading criteria might be.
4. Bring students back together and remove any redundancies in the grading criteria.
5. Once you have the grading criteria, you can choose to continue the rubric development with your students or without them.

1. If you choose to continue with students, then ask students to determine the weight of each criterion for the assessment.
2. Next, you can break students into groups and have each group describe a different grading criterion at a set number of performance levels (e.g., A, B, C, D).
3. Collect all the descriptions and create one analytic rubric from each group’s descriptions.
4. Ask students to review, check for mutually exclusive language, and discuss any changes needed as a class.
5. Utilize the rubric tool in Blackboard to build a clickable rubric for grading and a viewable rubric for students.
6. Implement the rubric (without making changes) for the entire term. Reflect on the use of the rubric and identify areas of improvement to make adjustments to criteria, descriptions, or weight for the next term.

For more information on building rubrics in your course site visit the Blackboard Grading and Assessments page on the CATE website to view the Getting Started with Rubrics section.

Criterion descriptions should not contain duplication of criteria within the description; in other words, you should not have two grading criteria that assess the same attribute or element (e.g., critical thinking or formatting, etc.). Each criterion should be specific without duplications in grading.

For example, you have created a checklist that has one criterion for showing work and another criterion for the correct answer. The correct answer criterion should only assess the correctness of the final answer, not the demonstration of the correct problem-solving in the work; this element should be assessed in the showing work criterion.

Adjectives and adverbs can be used to help describe the grading criteria at different performance levels but should be mutually exclusive. Mutually exclusive language means that an adjective used to describe the performance at the highest level shouldn’t be used to describe the performance at the next level. For example, you have used the adjective “thorough” to describe the level of details provided at the exemplary level. So, you should not use the same adjective to describe the proficient level of performance or the subsequent level.

Please note that the following list is not all-encompassing and should be viewed as a starting point for describing grading criteria.

• 100% – A level of performance could be described with any of the following terms: Exemplary, outstanding, distinguished, exceptional, well-developed, excellent, comprehensive, thorough, robust, expert, extensive, etc.
• 80% –  B level of performance could be described with any of the following terms: Proficient, above average, accurate, complete, skillful, accomplished, clear, concise, consistent, etc.
• 70% – C level of performance could be described with any of the following terms: Satisfactory, competent, average, adequate, reasonable, acceptable, basic, sufficient, etc.
• 60% – D level of performance could be described with any of the following terms: Developing, attempted, emerging, approaching, novice, partial, etc.
• 50% – F level of performance could be described with any of the following terms: Beginning, rudimentary, rarely, seldom, needs revision, no evidence, etc.

It is important to be consistent with the use of adjectives when developing a rubric or checklist. This consistency will help support student understanding of expectations as well as improve inter-rater reliability if more than one instructor or TA is utilizing the rubric for grading and scoring.

As you begin describing criteria, make sure to focus on the tangible items that can be more objectively measured. For example, if there is a grading criterion for the overall quality of the work, avoid adding subjective elements like “effort.” A student who has not developed the skills yet to perform highly on the assessment might have put in a lot of effort but may have still performed poorly.

Try to use supportive language when describing criteria that help instill a growth mindset in students. And try to avoid negative language that may demotivate students. For example, instead of describing a criterion as “lacking” an element, use the word “missing, developing, or beginning.” Also, consider using terminology like “attempted” at the C or D level; this helps recognize students’ efforts.

Lastly, when describing the grading criteria focus on the quality of the work. Utilize descriptions that help highlight the work’s quality and focus less on quantifying the students’ work. For example, if you have a grading criterion for the mechanics of writing, you can describe it without counting errors in a paper.

• The Exemplary level of performance could be described as “professional language used in a 2-page report with minimal to no errors in spelling, grammar, punctuation, and capitalization.”
• The Proficient level of performance could be described as “professional language used in a 2-page report with minor errors in spelling, grammar, punctuation, and capitalization (e.g., misplaced punctuation, homophone errors – to, too, two).”
• The Satisfactory level of performance could be described as “professional language used in a 2-page report with errors in spelling, grammar, punctuation, and capitalization (e.g., capitalization errors, missing punctuation, grammar, etc.) but still able to understand the point of view.”
• The Developing level of performance could be described as “attempted professional language in a 2-page report with numerous errors in spelling, grammar, punctuation, and capitalization that distract and cause unreadability.”

If creating an analytic rubric or a holistic rubric, it is recommended to start with the description for the highest level of performance (exemplary level). This level would typically receive an A percentage or point value.

Once you describe the highest level of performance then you can focus on the next level and then the next level, etc. Once you have described all the criteria for the rubric, make sure to check that you are not duplicating criteria and have mutually exclusive language.

• Exemplars can be former students’ work (with permission), published work (with permission), or instructor work.
• Present features of the exemplar during a class session and deconstruct the rubric using the exemplar to illustrate what the rubric descriptions mean.

• Think of exemplars as a guide for students to know how to start. Students will understand the expectations of the structure, style, layout, content, etc.
• Have students use the exemplar and rubric to self-assess their own work. Students will develop the ability to analyze their work to determine strengths and weaknesses and the ability to know how to make it better.

Consider the following questions to improve the validity and reliability of the assessment as you develop and review your rubrics (Lin, 2020):

• Does the rubric measure the learning objective(s) adequately?
• Does the rubric include aspects that are irrelevant to the learning objective(s) and/or task?
• Do the descriptions for grading criteria contain tangible, supportive, and qualitative terms?

• Does the rubric include any aspects that could potentially reflect assessment biases?
• Are the grading criteria distinct from one another and use mutually exclusive language?
• Are the grading criteria weighted appropriately?
• Are the levels of performance weighted appropriately?
• Is the rubric paired with an exemplar (exemplary work sample) to support students and multiple instructors’ understanding of expectations?

• Single-Point Rubric Template
• Single-Point Rubric for Music Performance

• Single-Point Rubric for an Authentic Assessment

• Analytic Rubric Template
• Analytic Rubric for a Presentation

• Analytic Rubric for Art
• Analytic Rubric for Group Work

• Holistic Rubric Template
• Holistic Rubric for Written Assignment

• Holistic Rubric for Discussion Participation
• Holistic Rubric for Essay Response

• Checklist Template
• Checklist for Computer Programming Assignment

• Checklist for a Math Assignment
• Checklist for a Science Report

Eberly Center. (n.d.). Grading and performance rubrics. Carnegie Mellon University.

Gonzalez, J. (2014). Know your terms: Holistic, analytic, and single-point rubrics. Cult of Pedagogy

Poorvu Center for Teaching and Learning. (n.d.). Creating and using rubrics. Yale University.

Teaching Commons. (n.d.). Rubrics. DePaul University.

Abdel-Magid, T., Abdel-Magid, I. (2020). Grading of an assessment rubric. 10.13140/RG.2.2.16717.38887.

Al-Ghazo, A., Ta’amneh, I. (2021). Evaluation and grading of students’ writing: Holistic and analytic scoring rubrics. Journal for the Study of English Linguistics. 9. 77. 10.5296/jsel.v9i1.19060.

Al-Salmani, F., Thacker, B. (2021). Rubric for assessing thinking skills in free-response exam problems. Physical Review Physics Education Research. 17. 10.1103/PhysRevPhysEducRes.17.010135.

Hazels, T., Schutte, K., McVay, S. (2020). Case study in using integrated rubrics in assessment. Journal of Education and Culture Studies. 4. p81. Doi: 10.22158/jecs.v4n3p81. http://dx.doi.org/10.22158/jecs.v4n3p81

Jescovitch, L., Scott, E., Cerchiara, J., Doherty, J., Wenderoth, M., Merrill, J., Urban-Lurain, M., Haudek, K. (2019). Deconstruction of holistic rubrics into analytic rubrics for large-scale assessments of students’ reasoning of complex science concepts. 

Lin, R. (2020). Rubrics for scoring, interpretations and decision-making. 10.4324/9780429022081-5.

Nkhoma, C., Nkhoma, M., Thomas, S., Le, N. (2020). The role of rubrics in learning and implementation of authentic assessment: A literature review. 237-276. 10.28945/4606.

Smyth, P., To, J., Carless, D. (2020). The interplay between exemplars and rubrics.

Tomas, C., Whitt, E., Lavelle-Hill, R., Severn, K. (2019). Modeling holistic marks with analytic rubrics.