rubric-designer

name	rubric-designer
description	Design assessment rubrics — analytic, holistic, or single-point — that match learning objectives and produce consistent, defensible grading.
version	1.0.0
author	Gnostor Community
license	MIT
metadata	{"hermes":{"tags":["education","assessment","rubric","grading","evaluation","teaching"]}}

Rubric Designer

Design rubrics that actually measure what you want students to learn — not just what's easy to count.

When to Use

• "Build a rubric for [assignment]"
• "I need to grade [project type] — design criteria"
• "Make this rubric clearer / more consistent"
• "Convert these learning objectives into a rubric"
• "Design a peer-review rubric"

What to Give Me

• Assignment: essay, presentation, project, lab report, portfolio, performance, code
• Learning objectives: what skills or knowledge does this assess?
• Grade level / context: K–12, undergrad, grad, professional
• Purpose: formative (feedback) or summative (grade)
• Stakes and time: high-stakes final vs. weekly check-in
• Length: how detailed should it be?

Rubric Types

Analytic Rubric

Each criterion scored separately. Best for detailed feedback and multi-dimensional work.

| Criterion | Exemplary (4) | Proficient (3) | Developing (2) | Beginning (1) |
|-----------|---------------|----------------|----------------|----------------|
| **Thesis & argument** | Original, nuanced claim defended throughout | Clear claim, mostly defended | Claim present but inconsistently supported | No clear claim or off-topic |
| **Evidence & analysis** | Compelling evidence; analysis goes beyond summary | Sufficient evidence; some analysis | Limited or weakly connected evidence | Missing or incorrect evidence |
| **Organization** | Logical flow; transitions reinforce argument | Generally organized; minor lapses | Some structural issues that distract | Hard to follow |
| **Mechanics** | Polished; errors do not impede meaning | Few errors; minor distractions | Frequent errors that affect clarity | Errors interfere with comprehension |

Holistic Rubric

Single score for overall quality. Faster to grade; less actionable feedback.

**4 — Exemplary**: [paragraph describing the gestalt of an A paper]

**3 — Proficient**: [paragraph for B-level]

**2 — Developing**: [paragraph for C-level]

**1 — Beginning**: [paragraph for D/F level]

Single-Point Rubric

One column describes proficiency; columns on either side are for "concerns" and "advanced." Excellent for formative feedback.

| Concerns | Criteria for Proficiency | Advanced |
|----------|--------------------------|----------|
| (notes on what fell short) | **Thesis is clear and defended** | (notes on what exceeded) |
| (notes on what fell short) | **Evidence is relevant and analyzed** | (notes on what exceeded) |

Checklist (when binary criteria suffice)

- [ ] Thesis stated in opening paragraph
- [ ] At least three sources cited
- [ ] Counterargument addressed
- [ ] APA formatting throughout

Design Principles

• Backwards from objectives — every criterion should map to a learning objective; if it doesn't, cut it
• Observable behavior — score what you can see in the work, not what you think the student knew
• Distinguishable levels — if you can't tell "3" from "4," the rubric isn't ready
• Avoid counting as a proxy for quality — "five citations" isn't the same as "appropriate use of evidence"
• Calibrate with examples — for each level, anchor with a sample student work if possible
• 3–5 levels typical — more granularity rarely improves consistency; less leaves you stuck between scores
• Weight criteria intentionally — if argument matters more than mechanics, the rubric should reflect that

Common Pitfalls

• Vague descriptors: "good," "well-written," "shows effort" — replace with specific, observable language
• Over-reliance on length or count: word count doesn't measure quality
• Unweighted analytic rubrics that treat trivial and central criteria equally
• Norm-referenced language: "better than peers" introduces inconsistency across cohorts
• Hidden criteria: never grade on something not in the rubric students saw

Output

I'll deliver:

• The rubric itself in your chosen format
• Notes on which criteria map to which learning objectives
• Suggested point distribution / weighting
• A short calibration guide if the rubric will be used by multiple graders
• Optional: rubric framed for the student (criteria as "I can..." statements)