Measurement is central to the practice of science, engineering and commerce, as well as many everyday activities. It is therefore no surprise that it has generated a substantial body of scholarship, which has examined its elements, conditions and limitations. Although most contemporary authors agree that measurement involves the associating of numbers with objects and phenomena, there is much disagreement over how to characterize what kinds of things can be measured, what exactly counts as a measurement, and what makes measurement possible.
Mathematical theories of what measurement is are concerned primarily with the mathematical properties of measurement scales, and the conditions that make them valid. These include ensuring that the units in a measurement are consistent, and that the measurements of the same object or phenomenon are coherent (for example, a person’s height may be reported in feet and inches, or in meters). They also examine the way in which a measurement is performed, including how it is interpreted, and whether it is sufficiently accurate and reliable to be useful.
More recent work has focused on the role of information in the measurement process, and how it relates to the accuracy of a measurement. This has been influenced by developments in information theory and communication, and the work of mathematicians such as Maxwell, Helmholtz and Mach in the nineteenth century, which analyzed the nature of physical quantities.
A common view is that a measurement involves determining the amount of information in a given state or set of states, and that this information can be used to represent the state. This information, in turn, can be used to construct a measurement scale, which can be used to compare different measurements of the same object or phenomenon. The resulting scales are then compared to each other to ensure that they are consistent.
However, this characterization does not adequately describe how measurement works, as it does not address what kind of information is required to construct a measurement scale, and how that information can be gathered. More recently, scholars have argued that a more precise description of the measurement process is needed. Model-based accounts of measurement describe the process as a complex interaction between an object of interest, an instrument, and an environment that includes the measuring subject. These interactions are represented in a model of the object’s system, and the aim is to locate the object on a particular region of the abstract parameter space, which reduces the range of possible values.
In this context, artifact-free definitions of measurement have been proposed that aim to define units without reference to a specific physical object that serves as a standard. These types of definitions are intended to be independent of the deterioration of the measurement artifact, and thus provide a more secure basis for unit construction. The most commonly used artifact-free measurement systems are the International System of Units (SI) and the metre, candela, second, ampere and kelvin.