Foundations · 4 of 40

Accuracy, precision, and uncertainty

Learn to ask how trustworthy a measurement is instead of judging it by how many digits it contains.

Why this matters

A phone, survey receiver, copied spreadsheet, and hand-entered location can all display many decimal places. The number of displayed digits does not tell you how close the original measurement is to the real position.

Learn it

Accuracy is closeness to the true value. Precision is how finely a value is represented or how consistently repeated measurements agree. Uncertainty describes the range of doubt around a result. Rounding a coordinate changes its stored precision; it does not repair an inaccurate measurement. Likewise, geohash length changes cell size, not sensor accuracy.

Work through an example

47.61 and 47.610000 carry different amounts of written precision, but adding zeroes does not create new knowledge about where the measurement came from.

Precision does not create accuracy

Stored value: 47.610000
Possible source accuracy: ±15 m
More digits ≠ better source accuracy

Practice

1Use the precision exercise to change decimal places and watch the represented position change.
2Imagine the original phone location was uncertain by 15 meters. Ask whether six decimal places removes that uncertainty.
3Write one sentence describing accuracy and one describing precision without using the words interchangeably.

Try it

How many digits are you storing?

Move the slider to change how many decimal places are shown. Watch the stored number change, then remember: this changes written precision, not the quality of the original measurement.

Decimal places: 5

47.60621

A coordinate can carry many digits and still come from a measurement with real uncertainty. Keep source accuracy or uncertainty information when you have it; do not infer accuracy from decimal places.

What to watch for

When a source provides an accuracy or uncertainty estimate, keep it. When it does not, do not invent one from decimal places.

Where this takes you

This matters in scientific observations, field work, asset placement, geofencing, navigation, telemetry, and any workflow where location quality affects a decision.