The usual definition of measurement is the assignment of numbers to something according to specified rules, where empirical evidence can be used to validate the results (Magnusson, 1966, see p.1).
But at least 40 different ways of using the term “measure” have been identified (Lorge, 1951). The same term can mean the act of weighing, the balance by which weighing is done, the grams that are used to balance an object, or the numeral that expresses the result of the weighing process. It can be used to refer to any instrument used as a basis for comparison, even when that comparison involves processes of estimation or judgment.
“Measure” is used more frequently to refer to acts of subjective estimates than to precise objective determination.
It is not likely that consensus will emerge regarding one meaning for “measure” or associated terms. Most people, however, expect some quantity will be used to express the outcome of some measurement process. In general, measurement involves assigning a class of numerals to a class of objects by applying a specific set of rules or procedures. (See Numbers and Numerals All measurement involves a person’s use of perceptual faculties, either unaided or extended by instrumentation of some sort.
Thus the classes of objects, the conceptual organization of discourse about those objects, the nature of instruments used, and the training of the observer all influence the results of a measurement process. Some observations are made directly — for example, the observer lays the ruler on the face of an object and compares the ends of the face with the markings on the ruler. The property of interest is observed directly.
Other observations are indirect — for example, the length of a column of mercury in a thermometer has been correlated with variations in temperature. In this case the effects of temperature on the column of mercury are observed, and then an inference is made about temperature by observing the column of mercury.
In scientific measurement, regardless of the field, the conditions for observation are carefully specified in terms of time, place, and circumstance. Observations are independently verified under the same conditions, with the results reported in terms of probable error.
Lorge (1951) makes the important observation that what is observed depends upon man’s conceptual equipment to translate sensory experiences into the notion of a property. Frequently the notion of a property will change as attempts are made to measure it. Detailed description of the property is essential.
Statements about a property are empirical in that they depend on what is experienced. Science demands that observations be reproducible — measurements must be reproducible. Assuming the property remains constant, measurement of the property under the same conditions by different observers should yield the same result.
Lorge, I. (1951). The fundamental nature of measurement. In E. F. Lindquist (ed.), Educational measurement. Washington, D.C.: American Council on Education. (pp. 533-559).
Magnusson, D. (1966). Test theory. Reading, MA: Addison-Wesley.