Compare and explore

AIR Research Laboratory

A governed workspace for asking new questions across Five Eyes social context and reviewed AIR case data without requiring statistical software or coding expertise.

Exploratory, not causal. Visual association, correlation, lag and temporal proximity do not establish why an incident occurred.

Compatibility before calculation. Units, coverage, age groups, source regimes, series breaks and missingness are screened using versioned permissions, structured break rules and the metadata available in the public release. This does not establish full pairwise equivalence.

Three comparison families

Begin with the relationship you want to examine

Active modes use versioned public data, transparent inclusion rules and source-qualified metadata. Inactive modes remain visible so the intended analytical architecture is clear.

A

Context against context

Compare compatible social, economic, digital, demographic, health, crime, migration, religion and trust indicators across countries and time.

Active · governed by indicator permissions

B

Context against AIR cases

Place reviewed annual AIR case measures beside compatible social-context indicators, with explicit common coverage and non-causal cautions.

Active · reviewed case data and governed context data

C

Case against case

Profile and cross-tabulate governed case dimensions, with linked filters, small-cell protection and exportable results.

Active · reviewed public case fields

Open exploratory workspace

Choose a comparison

Unavailable or qualified comparisons are explained. The engine does not calculate correlation coefficients or causal effects.

Reviewed case release available. Context–case and case–case comparisons use the 346 public v4.2.3 records and only the approved public fields. Blank values remain missing, sensitive small cells are suppressed, and all comparisons are descriptive and exploratory.

Comparison family
Countries
Select one or more; comparison permissions still apply.
Temporal context Contextual heuristics; never causal breakpoints.

Loading indicator comparison permissions…

Comparison output

Loading governed releases…

Checking

Preparing governed context indicators, temporal overlays and comparison permissions.

Common coverage
Aligned observations
Missing combinations
Releases

Missingness categories will appear after the comparison is prepared.

Accessible comparison table

Variables and sources

Active specification

Mandatory interpretation boundary: visual association, correlation, a shared trend, an era marker or temporal proximity does not establish that a contextual condition, platform or algorithm caused an individual incident.

Interpretive boundaries

What the laboratory can—and cannot—do

Exploration is designed to generate observations and testable questions, not to convert national trends into individual causal explanations.

It can
  • align compatible annual series;
  • show counts, shares, rates and indexed trends when valid;
  • expose missingness and source breaks;
  • produce reproducible descriptive views.
It cannot
  • infer individual motive from national indicators;
  • repair incompatible definitions by visual design;
  • treat a correlation as a mechanism;
  • make incomplete public-source records complete.

Read the methodology and limitations