How to create an assessment rubric: step by step

Why intuition-based grading does not work

When a teacher returns a piece of work with a grade and no explicit criteria, several things happen at once. The student does not know exactly what they did well or what they should improve. The teacher cannot explain consistently why that work is a 7 and not a 6. And two teachers assessing the same work with the same implicit criteria almost always arrive at different grades.

This problem has a name in assessment literature: evaluative subjectivity. It does not mean the teacher’s judgment is incorrect — it means that without explicit criteria, the same work can receive very different evaluations depending on who reviews it, at what time of day, or whether it is the first piece reviewed or the twentieth.

Rubrics solve this problem. A rubric is an assessment tool that explicitly describes quality criteria and the different performance levels for each criterion. It is not a bureaucratic form — it is an assessment contract: it tells students exactly what is expected before they begin working.

What a rubric is (and is not)

A rubric has three basic components:

What differentiates a good rubric from a list of scores is the descriptor. A useful descriptor does not say “the content is very good” — it says “the student explains the central concept accurately, uses at least two relevant examples and connects the topic to the context studied in class.” With that descriptor, the student knows what they need to do and the teacher knows what they need to observe.

What a rubric is not: a checklist (did they do X? yes/no), a score scale without description (content: 0-10 points), or an instrument designed after assessing to justify already-assigned grades.

Two types of rubrics by use

Holistic rubric: assesses performance as a whole, assigning a general level. It is faster to design and apply, but offers less specific information to the student.

Analytic rubric: assesses each criterion separately. It is more work to design, but offers much more precise feedback. It is the recommended type for complex work or situations where the student needs to know exactly what to improve.

	Holistic rubric	Analytic rubric
Design time	Short	Moderate
Grading time	Short	Moderate-long
Feedback detail	Low	High
Useful for	First impressions, quick assessments	Process work, projects, writing
Usefulness for student	General orientation	Detailed guide for improvement

For most situations where the rubric should guide learning (not just grading), the analytic rubric is more effective.

How to design a rubric in 5 steps

Step 1: Identify what is being assessed (not how)

The first step is to decide on the criteria. A common mistake is including criteria that measure effort or behavior (like “submitted on time” or “worked in class”) rather than learning. The criteria in a rubric must answer the question: what does the student need to demonstrate to show they learned what I wanted them to learn?

For an oral presentation on a scientific topic, relevant criteria might be: conceptual mastery, use of evidence, clarity of explanation, ability to answer questions. For an argumentative essay: thesis, argumentation, use of sources, textual coherence.

A reasonable number of criteria is between 3 and 6. With fewer than 3, the assessment is too global. With more than 6, the rubric becomes unmanageable for the teacher and overwhelming for the student.

Step 2: Define the performance levels

The levels can have descriptive names (excellent, satisfactory, developing, insufficient) or numerical ones (4, 3, 2, 1). Research by Susan Brookhart (2013) suggests that descriptive names are more useful than numerical ones because they carry information: “satisfactory” tells the student they met the objective; “developing” tells them they are on the way but did not reach it.

Four levels is the most common and most useful number: it allows distinguishing excellent work from acceptable and insufficient from that which is in process. Three levels sometimes collapse important distinctions; five or more make it very difficult to write descriptors that differ with clarity.

Step 3: Write the highest-level descriptors first

The highest performance level describes what it means to master the criterion. It is the reference point from which the other levels are defined. An effective way to write it: “The student [does X] [with what precision/depth] [under what conditions].”

Example for the criterion “use of evidence” at excellent level: “Uses at least three reliable sources, cites specific data or examples to support each main argument and explains the connection between the evidence and the conclusion.”

Step 4: Write descriptors for intermediate and minimum levels

Once the highest level is defined, the lower levels are built as progressively incomplete or partial versions of that performance. The minimum level does not describe what the student “did not do” — it describes what they did, even if insufficient.

Continuing the example: satisfactory level: “Uses at least two sources and cites evidence for most arguments, although the connection between evidence and conclusion is not always explicit.” Developing level: “Mentions sources but evidence is scarce or insufficiently specific.” Insufficient: “Does not use external sources or uses them without relation to the arguments.”

Step 5: Calibrate the rubric with work samples

Before using the rubric for assessment, it is worth applying it to some sample works (from previous years or anonymized) and verifying that the descriptors produce consistent evaluations. If two teachers applying the same rubric to the same work arrive at very different results, the descriptors need more precision.

How to use the rubric to improve learning

A rubric used only for final grading has half its potential. Its greatest value is in what it can do before and during the process:

Before starting the work: share the rubric with students when presenting the task. Read it together, discuss what each criterion means with concrete examples and resolve doubts. This eliminates the anxiety of “I don’t know what’s expected” and gives the student a clear guide for planning their work.

During the process: use the rubric as a self-assessment tool midway through. “Before submitting the draft, read each criterion and mark what level you think you’re at. Justify with evidence from the work.” This practice activates metacognition and usually produces more substantive revisions than the generic comment “review it carefully before submitting.”

For peer feedback: in the work gallery, students browse their classmates’ work with the rubric in hand. Peer assessment without explicit criteria usually produces superficial feedback (“I liked it,” “it was good”). With a rubric, the student has to specify which criterion and at what level they place the peer’s work, and why.

Involving students in rubric design

An advanced but very effective practice is to co-construct the rubric with students. Before designing it, the teacher can use the collaborative ideas wall to ask: “What do you think characterizes an excellent oral presentation?” or “How would you know your argumentative essay is quality?”

The responses reveal which criteria are comprehensible for the group and which need to be taught. A student who can describe what an excellent presentation is before making one has a much clearer idea of what they need to achieve. Co-construction also increases commitment to the criteria: when students helped define “excellent,” they understand that level as a genuine goal rather than an external imposition.

Free tools for creating and using rubrics

Rubric Generator — for designing it in minutes

The rubric generator allows defining criteria, levels and descriptors without starting from scratch. It generates a table ready to share with students. The teacher can use it as is or adjust it for the specific work. Sharing it from the first day of the task is the simplest change with the greatest impact on the quality of submitted work.

Work Gallery — for peer assessment with criteria

In the work gallery, students browse their classmates’ productions with the rubric in hand. Peer assessment without explicit criteria produces superficial feedback (“I liked it,” “it was good”). With a rubric, the student has to specify which criterion and at what level they place the peer’s work — and justify it. This develops critical thinking in addition to improving the assessment process.

Exit Ticket — for verifying they understand the criteria

Before students start the assessed work, the exit ticket allows verifying whether they understood what is expected of each criterion. A question like “Explain in your own words what ‘excellent’ means in the argumentation criterion” reveals misunderstandings that the teacher can resolve before the work is already advanced.

Collaborative Ideas Wall — for co-constructing the criteria

The collaborative ideas wall allows students to express what they believe “excellent” means before seeing the rubric. Negotiating that definition with the group is part of the learning: when students helped define the criteria, they understand them as a genuine goal rather than an external imposition.

Common mistakes

Vague criteria. “Creativity: high/medium/low” tells the student nothing. What does high creativity mean in this specific work? If the descriptor does not allow distinguishing two real pieces of work, it needs to be more specific.

Too many criteria. A rubric with 10 criteria produces more exhaustive assessments in theory, but in practice neither the teacher nor the student can manage it well. Choosing the 4 or 5 most important criteria and describing them precisely is more useful than covering everything superficially.

Sharing it only at the end. A rubric shared after submitting the work only serves to justify the grade. Sharing it before is what converts it into a learning tool.

Criteria that measure the same thing. “Clarity” and “organization” can overlap. If two criteria produce identical evaluations for the same work, they are probably measuring the same dimension with different names.

What competencies it develops

The rubric is not only an assessment tool for the teacher — it is a learning tool for the student:

• When the student uses the rubric to self-assess their draft before submitting, they are exercising self-regulation: monitoring their own performance and adjusting it based on explicit criteria.
• When they reflect on the gap between where they are and where they want to be according to the criteria, they are practicing metacognition: thinking about the quality of their own work.
• When they analyze what they need to improve to move from the “developing” level to the “satisfactory” level, they are developing the competency of learning to learn: using information about their performance to plan their improvement.
• When they assess a peer’s work with explicit criteria, they are exercising critical thinking: making judgments based on specific evidence.

How to start