Uncertainty is part of measurement

A measurement is more useful when it states how uncertain it is and where that uncertainty comes from.

Measurement uncertainty describes the dispersion of values reasonably attributable to a measured quantity. Repeated observations can reveal some components; calibration, instrument limits, and modeling assumptions can contribute others. More decimal places do not necessarily mean more information.

For example: A scale that repeatedly reports the same weight can still be miscalibrated. Repeatability alone does not establish accuracy.

Keep in mind: Uncertainty is not the same as a known error. A stated interval only means what its construction and assumptions support.

Think about it

Which uncertainty would remain in a result even if you repeated the measurement a thousand times?

Connected ideas

Falsifiability and its limits · The map is not the territory · Bayes rule and belief updates

Sources and further reading

• NIST Technical Note 1297: Guidelines for Evaluating and Expressing Measurement Uncertainty

NoteGuild starter collection · Original educational summary; source authors are not represented as platform members.