Chapter 4: Measurement and Units of Analysis
You should now have an idea about how to assess the quality of your measures. But measurement is a complex process, and some concepts are more complex than others. Measuring a person’s political party affiliation, for example, is less complex than measuring her or his sense of alienation. In this section we will consider some of these complexities in measurement. First, we will examine the various levels of measurement that exist, and then we will consider a couple of strategies for capturing the complexities of the concepts we wish to measure.
Levels of measurement
When social scientists measure concepts, they sometimes use the language of variables and attributes. A variable refers to a grouping of several characteristics. Attributes are those characteristics. A variable’s attributes determine its level of measurement. There are four possible levels of measurement; they are nominal, ordinal, interval, and ratio.
At the nominal level of measurement, variable attributes meet the criteria of exhaustiveness and mutual exclusivity. This is the most basic level of measurement. Relationship status, gender, race, political party affiliation, and religious affiliation are all examples of nominal-level variables. For example, to measure relationship status, we might ask respondents to tell us if they are currently partnered or single. These two attributes pretty much exhaust the possibilities for relationship status (i.e., everyone is always one or the other of these), and it is not possible for a person to simultaneous occupy more than one of these statuses (e.g., if you are single, you cannot also be partnered). Therefore, this measure of relationship status meets the criteria that nominal-level attributes must be exhaustive and mutually exclusive. One unique feature of nominal-level measures is that they cannot be mathematically quantified. We cannot say, for example, that being partnered has more or less quantifiable value than being single (note we are not talking here about the economic impact of one’s relationship status— we are talking only about relationship status on its own, not in relation to other variables).
Unlike nominal-level measures, attributes at the ordinal level can be rank ordered, though we cannot calculate a mathematical distance between those attributes. We can simply say that one attribute of an ordinal-level variable is more or less than another attribute. Ordinal-level attributes are also exhaustive and mutually exclusive, as with nominal-level variables. Examples of ordinal-level measures include social class, degree of support for policy initiatives, television program rankings, and prejudice. Thus, while we can say that one person’s support for some public policy may be more or less than his neighbour’s level of support, we cannot say exactly how much more or less.
At the interval level, measures meet all the criteria of the two preceding levels, plus the distance between attributes is known to be equal. IQ scores are interval level, as are temperatures. Interval-level variables are not particularly common in social science research, but their defining characteristic is that we can say how much more or less one attribute differs from another. We cannot, however, say with certainty what the ratio of one attribute is in comparison to another. For example, it would not make sense to say that 50 degrees is half as hot as 100 degrees.
Finally, at the ratio level, attributes are mutually exclusive and exhaustive, attributes can be rank ordered, the distance between attributes is equal, and attributes have a true zero point. With these variables, we can say what the ratio of one attribute is in comparison to another. Examples of ratio-level variables include age and years of education. We know, for example, that a person who is 12 years old is twice as old as someone who is six years old.