WHAT ARE MEASURES?

Measures are analogous to the hierarchy of data:

Data are the basis of a measurement program. In general, measures have some:

•Quantitative elements (a number—e.g., 42, 127)

•Qualitative elements (e.g., high, heavier, warmer)

•Unit-defined elements (e.g., degrees Fahrenheit, inches, pounds)

•Unit-less elements (e.g., a weighting factor).

To be rigorous, consideration should be directed toward several properties of measures in evolving effective measurement. The key ones are precision, accuracy, and validity:

•Precision is the degree to which a measure exactly expresses the property being measured; e.g., inches are a more precise measure of length than are feet.

•Accuracy is the property of a measure that describes the extent to which the measure value represents the property value being measured. The measurement method often determines accuracy. “Stepping-off” a distance is generally less accurate than using a measuring tape; however, the way the measure is taken also affects its accuracy.

•Validity is an evaluation of a measure’s capability to express the property being measured. For example, in software engineering projects, lines of code are not a valid measure of software size where size is related to the amount of work required to develop, maintain, or support that amount of software product.

Some other properties of measures that should be considered are whether the measure is direct or indirect. For example, a direct measure of an object’s weight is the reading from the scale on which it is placed. To obtain the weight of that same object, the indirect measure would be obtained by measuring its volume and multiplying that result by the density of the object.

In addition, measures are hard or soft. Hard measures are based on some physical or determinative property of what is being measured, whereas soft measures are based on opinion or perceptions. For example, today’s temperature compared to yesterday’s would be a hard measure that would tell if today was warmer, cooler, or the same. If several people were asked if they thought today was warmer, cooler, or about the same as yesterday, their answers would be a soft measure of the comparative temperature.