Nominal scale is the first and most immature type of the scales of measurement, as it satisfies only one property of the scales i.e. different items or values on this scale have unique identities, but they have no specific order and can be put in any order. Moreover, they have no numerical meanings. They are just different entities which can be identified by their labels, codes or names. For example Gender is a nominal scale which may either be a male or a female. Religious affiliation is another example of nominal scales. A person may either be a Christian, a Muslim, a Jewish, a Buddhist or a Hindu etc. When using a nominal scale we usually code each item of the scale with identical numbers to make them simple for further operations. For example we can use 1 for male and 2 for female and vice versa, when dealing with gender. Similarly we can use 1 for Christian, 2 for Muslim, 3 for Jewish, 4 for Buddhist and 5 for Hindu when dealing with religion. Here the numbers have no quantitative importance; they are merely used for identification purposes. Other examples of nominal scale include nationality, ethnicity, languages, individuals’ preferences for various things, national security numbers, national ID cards numbers, national passport numbers, model or serial numbers of various consumer products, vehicles' registration numbers and landline or mobile phone numbers etc. Basic mathematical operations like addition, subtraction, multiplication and division on nominal scales have no meaning. The main problem with this type of scales is that we can't apply most of the powerful and sophisticated statistical tools on it. Of the three measures of central tendency, Mean, Median and Mode only Mode can be calculated for nominal scales, since Mean and Median have no sense for this scale. Similarly measures of dispersion like standard deviation and variance are also useless for nominal scale. You should compare this scale with the other three types of scales: ordinal, interval and ratio, to better understand it.
