A variable can be measured several different ways. This variety in variable measurement is broken down into four levels. These levels are nominal, ordinal, interval, and ratio. In this blog I will talk about nominal and ordinal and I will address interval and ratio in the next post.
Nominal data is data that is broken into separate and discrete categories. The categories are mutually exclusive which means that no data can be in more than one category. Nominal data is also exhaustive in that all the data must go into one of the categories. This is one of the weakest forms of measurement because differences within the category cannot be account for because all data is forced to conform to category. Examples of nominal measurement would be gender because everyone who responds must be placed in one category or the other and there is no way for someone to be half male half female when using nominal classifications.
Ordinal measurement is used for ranking data. At this level, data is still nominal but the order matters. An example would be class standing which is freshman, sophomore, junior, and senior. The data is nominal in that there are categories but the order matters as a senior is a higher level in comparison to a freshman. There is still no attempt to differentiate within categories which weakens this level of measurement.
What level of measurement to use is dependent on what your research questions are. Research is guided by the question you ask.