The field of statistics recognises several different types of data, including nominal, ordinal, interval and ratio. Of these types, nominal is the least numeric and the most like categorisation. In fact, a synonym for "nominal" in this context is "categorical."
According to the "Dictionary of Statistics and Methodology" by W. Paul Vogt, a nominal variable is one that "distinguishes among subjects by sorting them into a limited number of categories, including type or kind." When a nominal variable is a unchangeable trait or characteristic of a research participant, the term "attribute variable" is also applicable.
Imagine a data table with test scores in one column. The remaining two columns indicate the gender of each test-taker and type of test administered. The test scores are not nominal data because they give specific numbers that can be added, averaged or placed in numeric order. Gender is a nominal variable since it simply categorises the test takers into a finite number of groups: male and female. The last column could be nominal or not. A nominal example would be to merely label each type of test, such as Form A and Form B. However, if the different types of test systematically vary in difficulty, the third column might instead hold ranked data, such as Grade 8, Grade 9 and Grade 10 reading levels. This is not considered nominal data because the variable can be meaningfully sorted into numeric order.
A nominal variable is a type of qualitative variable, which merely describes data rather than quantitatively measures it. Another type of qualitative variable is the ordinal variable or rank variable. An ordinal variable, such as a grade-based reading level, places data in a meaningful order, but the distance between ranks is non-quantitative. For example, Grade 8 reading level is not necessarily twice as difficult as Grade 4.
In contrast, quantitative variables measure rather than describe data. Interval variables are quantitative variables that lack a logical zero point. Temperature scales are a good example, where the Fahrenheit and Celsius scales have different, arbitrary zero points. Finally, ratio variables are quantitative variables with an inherent zero point that indicates total lack of the given property.
Statistics and Coding
Nominal data have limitations with respect to the types of statistical tests that can be applied to them. It makes no sense, for example, to average a nominal variable. This is a pertinent point because nominal data can be coded using numbers. For example, "1" in a data table might stand for "male," while "2" stands for "female." Note, however, that this coding scheme does not mean that two males equal one female. Using such a coding scheme, it is possible to inadvertently subject nominal data to an inappropriate statistical test. Doing so returns meaningless results. There are certain statistical tests, however, that can be used with nominal data. These include frequency tables, Chi-Square testing and Analysis of Variance, among others.
Etymology and Other Uses
The word "nominal" comes from Latin nominalis, "of or belonging to a name." In this respect, a nominal variable is literally one which names or labels. Since the Latin word for "name" was also used to refer to nouns, linguists use the term "nominal" as the adjectival form of the word "noun," just as "verbal" is the adjectival form of "verb." The term is also commonly used to indicate something that is minimal, insignificant or "in name only."
- 20 of the funniest online reviews ever
- 14 Biggest lies people tell in online dating sites
- Hilarious things Google thinks you're trying to search for