What Is a Metric?

Now that we’ve discussed how we might collect data as well as the types of variables for which we may collect data, let’s now discuss metrics.

How would you define a metric? Once you have a definition, check your answer.

A metric is a measurement of an underlying concept.

Some of our data that we collect may be utilized in their raw form as metrics. In other instances, variables may need to be combined to form meaningful metrics.

Analysts often need to measure various quantities as a means of tracking performance over time or comparing figures across units. You are likely used to measuring the physical dimensions of objects—for example, weights. These tangible variables present no challenges for measurement because you can observe the relevant quantities. But for intangible variables, such as self-esteem, optimism, or innovation, the challenges of measurement are much higher. For intangible variables such as these, the measurement process must ensure that what you are measuring is truly the underlying concept you want to quantify.

Is It a Metric?

There are three requirements for a measurement to be considered a metric:

First, anything measurable has to be expressed in the form of a number. If something cannot be reduced to a number, it cannot be measured.

Second, anything measurable needs to be expressed in terms of units that all people can understand with or without explanation. For example, say that a person's weight is 158 pounds. Pounds are easily understood by all people because they are part of a standard system of weights and measures. Not all units are so easy to understand. When you use a ratings point as a unit of measurement for television audiences, it requires an explanation. Such units used in specialized fields also need to be understood by the practitioners within that field.

The third requirement for a metric is that it should be based on a well-tested and widely accepted methodology of measurement. The method of measurement is important because it affects the quality of all the numbers that are generated throughout the process. For example, when you go into a doctor's office, your blood pressure is measured via an instrument that needs to be used in a specific way. If the nurse does not follow the exact procedure, perhaps by not tightening the cuff around your arm correctly or by not waiting long enough to take a reading, then the results will not be accurate. In the case of media measurements, there are well-understood procedures that need to be followed for any measurement to be taken correctly.