nhouseholder/design-critique

Design Critique — Quantitative UX Evaluation

Structured design critique framework producing a quantitative UX score (/40), cognitive load assessment, and persona-based testing. Adapted from pbakaus/impeccable critique system.

When This Fires

• During /site-review — as the design/UX evaluation component
• During /site-redesign Phase 7 (Visual QA) — to score the redesign
• When user asks to "critique", "evaluate", or "score" a design
• When /qa-test finds UX issues that need deeper analysis

Step 1: Nielsen's Heuristics Score

Read reference/heuristics-scoring.md for the full scoring guide.

Score each of 10 heuristics 0-4. Be honest — a 4 means genuinely excellent.

| # | Heuristic | Score (0-4) | |---|-----------|-------------| | 1 | Visibility of System Status | | | 2 | Match Between System and Real World | | | 3 | User Control and Freedom | | | 4 | Consistency and Standards | | | 5 | Error Prevention | | | 6 | Recognition Rather Than Recall | | | 7 | Flexibility and Efficiency of Use | | | 8 | Aesthetic and Minimalist Design | | | 9 | Help Users Recover from Errors | | | 10 | Help and Documentation | | | | TOTAL | /40 |

Rating scale: | Score | Rating | Action | |-------|--------|--------| | 36-40 | Excellent | Minor polish only | | 28-35 | Good | Address weak areas | | 20-27 | Acceptable | Significant improvements needed | | 12-19 | Poor | Major UX overhaul required | | 0-11 | Critical | Redesign needed |

Tag each issue found with severity: P0 (blocking), P1 (major), P2 (minor), P3 (polish).

Step 2: Cognitive Load Assessment

Read reference/cognitive-load.md for the full checklist.

Evaluate 8 items — each pass/fail:

• [ ] Single focus: Primary task completable without distraction from competing elements?
• [ ] Chunking: Information in digestible groups (4 items per group max)?
• [ ] Grouping: Related items visually grouped (proximity, borders, shared background)?
• [ ] Visual hierarchy: Immediately clear what's most important?
• [ ] One thing at a time: User can focus on a single decision before the next?
• [ ] Minimal choices: 4 or fewer visible options at each decision point?
• [ ] Working memory: User doesn't need to remember info from a previous screen?
• [ ] Progressive disclosure: Complexity revealed only when needed?

Scoring: 0-1 failures = low load (good). 2-3 = moderate (address soon). 4+ = high load (critical fix).

Working Memory Rule (Cowan, 2001): At any decision point, count distinct items user must simultaneously consider. 4 or fewer = manageable. 5-7 = group or disclose progressively. 8+ = overloaded.

Step 3: Persona Testing

Read reference/personas.md for full persona profiles and red flag checklists.

Select 2-3 personas most relevant to the interface type:

| Interface Type | Use Personas | |---|---| | Landing page / marketing | Jordan (first-timer), Riley (stress tester), Casey (mobile) | | Dashboard / admin | Alex (power user), Sam (accessibility) | | E-commerce / checkout | Casey (mobile), Riley (stress tester), Jordan (first-timer) | | Onboarding flow | Jordan (first-timer), Casey (mobile) | | Data-heavy / analytics | Alex (power user), Sam (accessibility) | | Form-heavy / wizard | Jordan (first-timer), Sam (accessibility), Casey (mobile) |

For each selected persona, walk through the primary user action and report specific red flags — not generic concerns.

The 5 personas:

1. Alex (Impatient Power User) — Skips onboarding, wants keyboard shortcuts, bulk actions
2. Jordan (Confused First-Timer) — Reads everything, hesitates, needs guidance at every step
3. Sam (Accessibility-Dependent) — Screen reader, keyboard-only, needs 4.5:1 contrast
4. Riley (Deliberate Stress Tester) — Tests edge cases, empty states, long strings, refresh mid-flow
5. Casey (Distracted Mobile User) — Thumb-only, interrupted mid-flow, slow connection

Output Format

DESIGN CRITIQUE
===============
Project: [name]
URL: [if applicable]
Date: [date]

HEURISTICS SCORE: [N]/40 — [Rating]
Weakest: [heuristic name] ([score]) — [why]
Strongest: [heuristic name] ([score])

COGNITIVE LOAD: [Low/Moderate/High] ([N]/8 passed)
Key violation: [worst offender]

PERSONA TESTING: [N] personas tested
- [Persona]: [pass/fail] — [key finding]
- [Persona]: [pass/fail] — [key finding]

TOP ISSUES (by severity):
P0: [list or "None"]
P1: [list]
P2: [list]
P3: [list]

RECOMMENDATION: [Ship / Fix P0-P1 first / Major rework needed]