SozoRockTechnology

Perspective / Data / September 9, 2026

Evidence needs
its context.

A number should not lose its meaning on the way to a decision.

Start with what was measured

A county estimate describes a defined population in a defined place. It cannot become a neighborhood measurement simply because a user searched for a neighborhood. The interface needs to preserve the measurement geography.

Keep time in view

A dataset’s release date and the years underlying its measurements answer different questions. Display both. A recently published release may incorporate observations from earlier years.

Carry uncertainty into comparisons

Point estimates can look more decisive than the evidence permits. Display confidence intervals alongside the values. A difference between two estimates is not, by itself, a test of statistical significance or evidence of causation.

Make inspection part of the experience

In the CB-CAP public preview, a reader can select a county, compare the same measure within a state and trace the selected estimate to its source context. Missing values remain unavailable rather than becoming zero.

Explore CB-CAP

Source

The preview uses CDC PLACES county data, 2025 release. Read the dataset’s methods and measure definitions before interpreting a result.