Saltar al contenido
AKOS
Todos los packs
StableUXnielsen-norman-group

Nielsen Norman Group — Usability Heuristics & Research Practice

Operationalizes the Nielsen Norman Group school: the ten usability heuristics as a working evaluation system, plus the research discipline behind them (heuristic evaluation, severity ratings, discount usability). This is AKOS's backbone checklist vocabulary — most UX findings cite one of the ten heuristics as their category.

usabilityheuristicsheuristic-evaluationseverityresearchdiscount-usability

Principios

  • · M1 — Heuristics are lenses, not laws. Two can conflict (H7 accelerators vs H8 minimalism); resolve with the decision framework and profile.
  • · M2 — Tag findings. Every UX finding cites its heuristic — the tag carries justification and speeds triage.
  • · M3 — Severity = frequency × impact × persistence. Rate every finding; unrated findings don't prioritize.

Patrones

  • · One heuristic at a time. Sweep the screen ten times with one question each, not once with ten questions. Blended passes miss H5 and H7 systematically.
  • · The status question first. "If I walked up to this screen mid-task, would I know what's happening?" — H1 violations are the cheapest to find and fix.
  • · Read every word out loud as the user. Jargon (H2) hides in labels the team stopped seeing years ago.
  • · Try to leave. From every modal, wizard, and flow: is there an obvious, safe exit (H3)? Try Escape, back, and the X — all three.
  • · Hunt twins. Find the same concept in two places; check word, placement, and pattern match (H4). One inconsistency predicts a family.

Riesgos

  • · Heuristics as decoration — running the checklist after decisions are frozen, findings filed unread. Evaluation must precede commitment.
  • · Severity inflation — everything HIGH; prioritization dies. Rate honestly, use the persistence multiplier.
  • · One-pass exhaustiveness claim — "we heuristic-evaluated it, it's done." Inspection samples; iterate (evaluator effect).
  • · Testing to confirm — recruiting friendly users, leading tasks, helping mid-test. Tests exist to surprise you.

Checklist

  • [ ] No roach motels: every enrollable thing is exitable; every flow has an exit. (H3, NG9, NG11)
  • [ ] Actionable errors cannot be missed (persist until handled) and never destroy input. (H9, NG31)
  • [ ] High-cost commitments restate their object and consequences before commit. (H5/H6, NG19, NG20)
  • [ ] All async operations show status; background work has a revisitable surface. (H1, NG1–NG4)
  • [ ] Modals close via X/Escape; multi-step flows have state-preserving Back. (H3, NG8, NG9)
  • [ ] One component set; identical actions labeled and placed identically. (H4, NG12, NG13)

Fuentes

En la bibliografía AKOS