Methodology / version 4.0

Read each indicator on its own terms.

Humain Alert displays separate indicators. It no longer calculates or publishes an overall score. Units, reference years and coverage remain visible.

What the dashboard shows

Three gauges visualize specific pressures: restrictions on expression, intentional homicide and suicide mortality. These are not comprehensive assessments of freedom, personal safety or psychological wellbeing. Two further contextual measures retain their own units: child labour worldwide and consumption-based carbon emissions.

IndicatorOriginal unitDisplay treatment
Expression restrictionsV-Dem freedom-of-expression index, 0–1(1 − index) × 100
Intentional homicideDeaths per 100,000 peoplemin(rate ÷ 30 × 100, 100)
Suicide mortalityDeaths per 100,000 peoplemin(rate ÷ 30 × 100, 100)
Child labour worldwidePercentage of children aged 5–17ILO–UNICEF global estimate: 7.8%, 2024. No gauge or conversion.
Consumption-based CO₂Tonnes per person per yearOriginal physical unit. No gauge or conversion.

Editorial scales, not official ratings

The gauge ceilings and colour bands are Humain Alert’s visual choices, not thresholds endorsed by the data publishers or a scientific diagnosis. A normalized value is not a percentage of people affected. A low homicide rate does not establish that a location is safe for a particular person; a suicide rate does not measure all aspects of mental health.

Gauge bands are: below 10 Stable; 10 to below 25 Fragile; 25 to below 40 Concerning; 40 to below 60 Critical; 60–100 Emergency. They describe positions on this editorial scale only. No weights, combined score or country ranking are calculated.

When a gauge is withheld

Each gauge input needs a finite, non-negative value and a known reference year no more than five calendar years before the current UTC year. For a range, its oldest year must meet that limit. Unknown, future or reversed year ranges are excluded. Expression restrictions must also lie between 0 and 100.

For a group of countries, at least 80% of the countries in that defined group must contribute observations. For example, an Africa group containing 54 countries needs at least 44 contributors; 4 of 54 is insufficient. A known national observation or an explicitly labelled official aggregate can also qualify. Unknown coverage does not qualify. These five-year and 80% limits are editorial safeguards; passing them does not prove representativeness.

When either test fails, the needle and alert colour disappear. The card explains why and links to the raw-data table. Raw observations remain visible with their reference years and coverage, without being passed off as current regional conclusions. Missing values never become zero.

Regional averages and the World entry

Country-group figures are equally weighted means of available national observations, not population-weighted prevalence. A small country contributes as much as a large country. Coverage and the oldest–newest reference years are shown, so this limitation can be assessed.

The World entry uses an official worldwide aggregate when one exists. Expression restrictions can use an explicitly labelled mean of the countries in the V-Dem source; its denominator describes that source’s observed countries, not every person worldwide. Other national values are not silently averaged to invent an official World figure.

Two child indicators must remain distinct

The 7.8% worldwide child-labour card uses the ILO–UNICEF estimate for ages 5–17 in 2024. It remains worldwide context when another region is selected and does not change after a World Bank refresh.

For other geographies, the raw-data table may show the World Bank children-in-employment series for ages 7–14 (SL.TLF.0714.ZS), often with much older survey years. That series has a different population and definition. It is identified separately, is not substituted for the global estimate and is not normalized into a child-protection gauge.

How the carbon footprint is defined

Consumption-based CO₂ from the Global Carbon Budget via Our World in Data adjusts territorial fossil-fuel and industrial emissions for carbon embedded in imports and exports. It excludes land-use change and does not allocate international aviation and shipping to individual countries. It is a national average, not a visitor’s personal footprint, and does not measure all greenhouse gases.

Published evidence and source checks

The reviewed HTML snapshot retains its original observation dates. The browser can check five public endpoints and cache results locally for 24 hours; a failed check may be retried after 15 minutes when the dashboard is active. Successful retrieval does not imply that the observations are new. Source failures retain the last available values and the eligibility rules still apply.

Figures on editorial profiles are dated snapshots, so they can differ from the latest browser retrieval. Compare units, definition, geography, coverage and year before comparing values.

Sources and corrections

Methods and factual errors are documented in the corrections log. Our editorial policy explains sources, AI assistance and accountability.