B3.12.1Severity Ratingdesignresearch

Severity synthesizes frequency, impact, and persistence

Aliases: severity · usability problem rating · usability severity

What it is

A severity rating cannot ask only whether users became stuck. A usability problem’s level should synthesize frequency, task impact, and persistence: an occasional obstacle on a secondary path differs from a defect encountered by every new user, causing data loss, or blocking the task entirely.

Why it happens

Frequency determines exposure in people and repetitions; impact measures task failure, time loss, errors, emotional cost, safety, and business consequence; persistence asks whether users can recover, whether workarounds recur, and whether learning removes the problem. None substitutes for the others: frequent minor friction can accumulate into major cost, while a rare unrecoverable error may be critical. Synthesis should record evidence separately, then apply a team-defined weighting to form a level.

Studying it

Aggregate exposure, failure rate, time cost, abandonment, consequences, and recovery from usability tests, logs, support tickets, and field observation. Create anchored descriptions—for example, 0 not a problem, 1 minor, 2 moderate, 3 important, 4 blocker—and have raters score independently while recording assumptions. Use calibration tasks to compare ordering stability across combinations of problems.

Where it stops holding

A fixed scale is not directly comparable across products because audience size, task criticality, and business constraints differ. Frequent low-impact friction may rank high in consumer products, while rare catastrophic issues must be immediate in safety-critical systems. Without exposure data, ratings contain uncertainty and should include evidence strength.

Applying it

  • Record frequency evidence, task impact, recovery, affected roles, and business constraints for every issue.
  • Define anchors and escalation rules: blockers, data loss, safety, and legal issues go straight to the highest level.
  • Combine the three factors with one template so ordering is not based merely on apparent seriousness.
  • Re-review before release: remeasure frequency and impact after fixes rather than retaining old scores.

Related

  • Same group: B3.12.2 Ratings support ordering rather than absolute judgment · B3.12.3 Multiple evaluators rate independently before merging
  • Nearby: Q2 Usability Evaluation · Q4 Research Methods and Evaluation
  • Search terms: severity rating · usability problem · prioritization

Cards in the same group

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/handbook/B3.12.1