Why a published number actually moved, after checking whether the counting changed first.
3 credits·in-depth·about 2–4 min·9 domains
What it answers
The question Why It Changed is built for
Name the metric and the move. It first asks whether the definition or the method of counting changed, then reads what moved it, checks whether comparable series moved too, records what the coverage says and what is not driving it, and names what would settle the question.
Who reaches for it
Analysts, reporters and officials who get asked "why did this number go up" and want a better answer than the first headline.
What comes back
8 sections, on every run
These are the headings the result renders, read from the same promise the form makes before you spend a credit. A section that depends on what the record holds is rendered when it can be and never promised here.
01
Did the Counting Change?
02
The Read
03
What Moved It
04
Did Comparable Series Move Too?
05
What the Coverage Says
06
What Is Not Driving It
07
What Would Settle It
08
Open Questions
What you give it
2 boxes, 1 required
The statistic to explain
required · up to 300 characters
The window or claim
optional · up to 600 characters
Try it on: a subject you already hold an opinion about, so you can grade the result against what you know. Each one opens the form filled, in the General domain.
Global piracy incidents reported to the IMBUK net migration in the ONS seriesGerman industrial production in the Destatis series
What it will not do
The lines it holds
It reads the move, it does not predict the next one. A cause is offered with the evidence for it; a popular explanation the data does not support is listed under what is not driving it.
Every claim carries its grade
A claim the scan could not source is marked as such rather than dressed up. How the grades work →
Verified or grounded
Reported or attributed
Contested
Open question
Where it runs
Available in 9 domains
A domain may name it in its own words and price it on its own terms; each row is what that domain shows.
The fastest way to judge the result is to pick a subject you know cold and read it against what you know. The scan names its credit cost before it runs.