{"id":2016,"slug":"rule-r5-output-tiers-not-scores","title":"R5 — Output tiers, not precise scores","kind":"rule","scope":"business","status":"current","audiences":["kevin","candid-team"],"topics":["self-report-validity","survey-question-design"],"reference_body":"**Rule:** The widget output is a small set of relative-difficulty tiers (~3-4). Never a numeric score, percentage, or precise dollar figure.\n\n**Why:** (a) Self-report error is irreducible — Parry 2021 meta-analysis ([[parry-2021-nature-meta-self-report-divergence]]) shows weak self-vs-logged correlations even with sincere effort. (b) The bias direction is upward (overplacement), so any precise score inherits the inflation. (c) Tiers are robust to residual bias in a way scores are not. (d) Tiers also gain the precision-without-false-precision benefits described in the presentation brief [[research-brief-smb-widget-presenting-tiers-june-2026]].\n\n**How to apply:** All widget outputs land in {Low / Moderate / High / Very High} difficulty tiers — never 47/100, never $X-Y, never percentile.","rationale_body":null,"metadata":null,"links":{"outgoing":[{"slug":"parry-2021-nature-meta-self-report-divergence","title":"Parry et al. 2021 — pre-registered meta-analysis: self-report ≠ logged media use","kind":"reference","scope":"business","link_type":"depends-on"}],"incoming":[{"slug":"research-brief-smb-widget-capture-layer-june-2026","title":"Research brief: SMB widget capture layer — what owners can vs cannot self-report (June 2026)","kind":"reference","scope":"business","link_type":"depends-on"}]},"created_at":"2026-06-23T19:16:01.069Z","updated_at":"2026-06-23T19:16:01.069Z"}