Column 2: evidence · the two layers
Open data, followed all the way down to one person inside it.
Every piece here starts with a public dataset and ends with a story. The numbers can be checked: any figure set in mono opens its own evidence in a column to the right, so you never have to leave the sentence you are reading to find out whether it is true.
open to work
work
files
Left: where each occupation ranks when its exposure is predicted from the 21 abilities it is built from. Right: where it ranks when the dataset rates the occupation whole. Same 271 jobs, two measurements. Orange lines rise, blue lines fall.
A correlation of −0.367 is not a weak relationship — it is an inverted one. The thick orange line is Writers and authors, moving from rank 267 to rank 4.
consistency test
largest disagreements, both directions
| occupation | predicted | rated | gap |
|---|---|---|---|
| Writers and authors | -0.057 | 0.774 | +0.831 |
| Public relations specialists | 0.020 | 0.788 | +0.768 |
| Models | -0.220 | 0.400 | +0.620 |
| Urban and regional planners | -0.036 | 0.580 | +0.616 |
| … | |||
| Automotive service technicians and mechanics | 0.757 | 0.073 | −0.684 |
| General maintenance and repair workers | 0.787 | 0.078 | −0.709 |
| Janitors and building cleaners | 0.746 | 0.028 | −0.718 |
Everything predicted too safe works through words. Everything predicted too exposed works through objects.