Methodology

How to read this dashboard

The Global Literature Dashboard mixes official statistics, trade journalism, and editorial analysis on a single screen. That mixture is the point — but it only works if every item tells you, plainly, what kind of claim it is. This page explains the labels you'll see throughout the dashboard and the editorial stance behind them.

Evidence tiers

Every signal, book, quote, event, and idea on this dashboard carries an evidence tier badge. The tier describes where a claim comes from, not how important it is. A Tier 3 editorial idea can matter more than a Tier 1 statistic — it is simply a different kind of claim, and you should weigh it differently.

Tier 1 · Official

Official data

Tier 1 items come from primary sources with an institutional stake in accuracy: ISBN registration agencies, UNESCO and WIPO statistical releases, national publishing and book-trade associations, and the official programs of festivals and prizes.

Examples on this dashboard: the Polish Book Institute's translation grant announcement, WIPO's AI Act policy briefing, and festival program pages for Frankfurt, Jaipur, and the Vancouver Writers Fest.

Tier 1 does not mean “always current.” An annual grant figure is Tier 1 and trustworthy, but it is still only as fresh as its last publication — see Freshness below.

Tier 2 · Reported

Reported

Tier 2 items come from reputable literary journalism and trade press: outlets like Publishers Weekly, LitHub, Asymptote, Words Without Borders, ArabLit, Brittle Paper, China Books Review, and Scroll.in Books. These outlets do original reporting and have editorial standards, but they are secondary sources — someone is reporting on a market, a deal, or a trend, not publishing a primary record of it.

Most of the day-to-day signal feed on this dashboard is Tier 2. That is by design: this is where most of the texture of the global literary world actually gets written down.

Tier 3 · Editorial

Editorial intelligence

Tier 3 items are this dashboard's own analysis: composite indices like Translation Momentum or Emerging Market Activity, and the Dangerous / Neglected Idea panel. They are built from Tier 1 and Tier 2 inputs but go beyond what those inputs strictly say — they interpret, model, or speculate.

Tier 3 content is visually distinguished throughout the dashboard (a warm accent border and an “Editorial” label) so it is never mistaken for a measured fact. If you find a Tier 3 claim styled like hard data anywhere on this dashboard, that is a bug, not a feature.

Confidence levels

Confidence describes how sure we are that a claim is accurate given its sources. It is independent of evidence tier and, critically, independent of importance.

High confidenceVerified by official data or multiple corroborating sources.

Medium confidenceReported by a credible source but not independently corroborated.

Low confidenceEarly, partial, or single-source reporting. Treat as directional.

Editorial confidenceAn in-house interpretation or hypothesis, not a measured fact.

Unknown confidenceConfidence has not been assessed for this item.

A useful test: the Dangerous / Neglected Idea panel is deliberately the most important item on the overview page and carries Editorial confidence. Confidence tells you how settled a claim is, not how much it deserves your attention. Treat a Low-confidence Tier 2 report about a market shift as worth watching even though it is less “certain” than a High-confidence Tier 1 statistic about something that already happened.

Freshness

Freshness describes how often a source updates — not how old the specific item you're looking at is. It exists to prevent a very specific failure mode: presenting an annual statistic with the same visual urgency as a live feed.

LiveUpdated in real time or near real time. Nothing on this prototype currently carries this label — it is reserved for future data sources that genuinely warrant it.

WeeklyRefreshed roughly once a week — typical for trade press coverage.

MonthlyRefreshed roughly once a month — typical for slower market reporting.

QuarterlyRefreshed a few times a year — typical for policy briefings.

AnnualRefreshed once a year — typical for grant programs and official statistics. Annual data is never presented with a “live” indicator, no matter how current it feels.

HistoricalA fixed historical reference point that will not update — used for the occasional “historic echo” quote.

The rule of thumb: an Annual, High-confidence, Tier 1 figure is trustworthy and slow-moving. A Weekly, Medium-confidence, Tier 2 signal is timely but provisional. Neither is “better” — they answer different questions.

Editorial stance

This dashboard is built on the premise that showing uncertainty is more trustworthy than hiding it. Most dashboards present everything in the same typography and color, which quietly implies that everything has the same epistemic status. It doesn't. A festival program, a translator's interview, and an in-house index are different kinds of claims, and treating them identically is itself a misleading editorial choice.

In practice, this means: every data-bearing card on this dashboard shows its source, evidence tier, confidence, and freshness, even when that makes the card busier. Tier 3 editorial content is visually marked so it can never be mistaken for hard data, even by a reader skimming quickly. And when an item is included because it is locally relevant — for example, items connected to Vancouver, the Lower Mainland, or British Columbia — it is flagged as such, but placed inside the same global feeds as everything else, because it earned its place by clarifying something, not by being local.

This is Phase 1 of the Global Literature Dashboard: a static prototype built on carefully written mocked data, designed so the information architecture and content model hold up once real data sources — and eventually live feeds — are connected behind them.