Writing Rubrics
Write effective rubrics with clear criteria for each rating level.
What is a Rubric?
A rubric defines what each rating level means for a given dimension. Clear rubrics help evaluators rate consistently and give Autousers precise criteria to evaluate against. Good rubrics are the foundation of reliable UX evaluations.
Writing Effective Rubrics
- Start with the extremes: define what a 1 (worst) and 5 (best) look like.
- Fill in the middle levels, ensuring each is clearly distinct from adjacent levels.
- Use specific, observable criteria rather than vague adjectives.
- Include examples where possible to illustrate each level.
- Test the rubric by having multiple people rate the same experience and checking for consistency.
Good vs Bad Rubric Criteria
Bad rubric criteria are vague, like "design is nice" or "somewhat usable." Good rubric criteria are specific and observable, like "primary call-to-action is immediately visible above the fold" or "form fields have clear labels and validation messages."
Rubrics improve both human and autouser consistency. When human evaluators disagree, it is often because the rubric criteria are ambiguous. Refining your rubrics is the single most effective way to improve evaluation quality.
Rubric Templates
Autousers provides rubric templates for common dimensions like Visual Design and Usability. You can use these as starting points and customize them for your specific evaluation context.
Was this article helpful?