Glossary

The value for your data. Usually the mean, but occasionally something else.

A quantity, like time or mass, whose values are continuous: they can take on any value with an infinite number of decimal places.

Variables, such as dice rolls or counting atoms, that can only take on specific values. You cannot have half an atom!

The indicated measurement is the observational result of a continuous variable as reported by your measuring device, which has a limited precision.

A property of the entire population. The average height of all the people on Earth would be a parameter. Parameters are often measured by statistics which are based off samples.

The uncertainty on a measurement divided by the mean. Generally a more useful quantity than percent error which requires a true value.

The complete set of persons, places, things, etc. understudy. Often this set is too large (it may even be infinite) to study in its entirety. As such, we usually sample.

Measuring time, for example, is a sample. You are doing a few measurements of the infinite number of possible values.

A subset of the population under study.

Measuring time, for example, is a sample. You are doing a few measurements of the infinite number of possible values.

A number that represents a property of the sample. Examples could be mean, median, or standard deviation.

Uncertainty arising from any measurement of a continuous variable. The true value is not only unknown, but unknowable due to its infinite number of digits. Moreover, there is intrinsic randomness from measurement to measurement.

A consistent difference between the indicated and true values, usually arising from a miscalibrated instrument or neglected effect. A systematic uncertainty is always in the same direction as opposed to the random bouncing around characteristic of statistical uncertainties.

The unknowable actual value of a continuous variable including the infinite number of decimal places.