MyLeoNes™

Measurement bias — Data, 14–17 years

Learn how a measuring method can push results in one direction. Data, 14–17 years.

What measurement bias is

Measurement bias happens when the way we collect a value tends to make it too high or too low. It is not just a random mistake: repeating the same unfair method can keep producing the same kind of error. A bathroom scale that always adds 2 kg is an example.

Why it matters

The problem is that a precise-looking number can still be wrong in a predictable way. Scientists and engineers began checking measurement methods because decisions about health, safety and money depend on them. A ruler with a worn-out zero point can make every length seem slightly larger.

Finding the bias

A scale is checked with a 10.0 kg weight. It reads 10.3 kg, so its bias is +0.3 kg. It then reads 6.8 kg for a parcel; subtracting 0.3 gives an estimated 6.5 kg. This correction is useful only if the scale’s error stays about the same.

The trap

A common mistake is to treat every difference from the truth as bias. That is reasonable because both bias and random error make one reading wrong. The key question is whether repeated readings lean in one direction; scattered readings may be random error instead.

Where it appears

Measurement bias matters when checking air pollution, body temperature, rainfall or the energy used by a device. A badly placed sensor can report a local condition rather than the wider situation. Calibrating equipment and comparing it with a trusted standard help reveal the problem.

Keep exploring

Other languages

Loading MyLeoNes™…