23 junio, 2024

Nominal variable: concept and examples

A nominal variable It is one that takes values ​​that identify a class or category in which the objects of study are grouped. For example, the variable ‘hair color’ groups people into brown hair, black hair, blonde hair, etc.

The nominal scale identifies, groups and differentiates the study units, according to a certain quality, in clearly defined and exclusive classes, in such a way that all those that belong to a class are equal or equivalent with respect to the attribute or property under study.

The classes are differentiated with names or with identifier numbers, so they do not have a numerical value or established order. For example: the variable sex has two classes, masculine and feminine; the numbers 1 and 2 representing the male and female categories, respectively, can also be used. These numbers are just arbitrary identifiers.

In this type of measurement, names or labels are assigned to objects. The name of most nominated specimens or definitions is the «value» assigned to the nominal measure of the object of study.

If two objects have the same name associated with them, then they belong to the same category, and that is the only meaning nominal measures have.

[toc]

Concept and characteristics

The nominal scale is the most elementary and the variables measured on this scale classify the study units (objects, people, etc.) into classes, based on one or more unique and observed characteristics, attributes or properties.

Classes or categories have a name or number, but these serve merely as labels or identifiers, make categorical rather than quantitative distinctions, serve a purely classification function

They cannot be manipulated arithmetically, they do not reflect order (ascending or descending) or hierarchy (higher or lower), the observations cannot be ordered from lower to higher or from small to large, that is, none of the categories has a higher hierarchy than the other. another, they are only reflecting differences in the variable.

Nominal variables with two classes are called dichotomous, such as the variable sex (male or female). Variables with three or more categories are called multichotomous or polychotomous. For example: the occupation variable (worker, carpenter, doctor, etc).

Nominal variables only establish equivalence relations; that is, a particular object of study either has the characteristic that defines the class or it does not.

With the nominal variables, calculations of proportions, percentages and ratios can be made, and with them frequency counts or tabulations of the number of events in each class of the variable studied are carried out. The measure of central tendency that can be handled with this type of variable is the mode.

Examples of nominal variables

Examples of variables measured on a nominal scale:

– Nationality (Argentine, Chilean, Colombian, Ecuadorian, Peruvian, etc.).

– Colors (white, yellow, blue, black, orange, etc.).

– Eye color (black, brown, blue, green, etc.).

– Classification of students by careers (Administration – 1; Systems – 2; Electronics – 3; Law – 4; etc.). (the number is a code with no value or order)

– Marital status (single, married, widowed, divorced, free union).

– Profession (engineer, lawyer, doctor, teacher, etc.).

– Sex (male, female).

– Religious affiliation (Christian, Muslim, Catholic, etc.).

– Political affiliation (liberal, conservative, independent, etc.).

– Type of school (public or private).

– Race (white, black, yellow, mestizo, etc.).

– Blood groups (O, A, B, AB).

– Examples explained

Soccer game attendees

If a count of attendees entering a football match is made, the nominal variable ‘attendance by gender’ can be defined. The count reports how many men and how many women attended the game, but the classification variable is gender.

Divide the crowd at the soccer game into two categories and neither group has preference over the other. Finally, the categories are exclusive, since there is no doubt to which group each of the attendees belongs.

Benefit of labor policies

It is desired to know the opinion of the people before the application of reforms in the labor policies of a country. The variable of ‘interest’ is the benefits of labor policies, and in the survey there are five possible positive results: More money, Better health care, Better retirement, Work/family balance and Other.

All responses are measured on a nominal scale with Yes or No values. The Other result encompasses all those benefits that respondents consider they will obtain, but that are not part of the survey values.

The number of affirmative or negative responses is necessary to calculate the percentage of respondents out of the total who consider that they will improve or not in any of the aspects, but these percentages have no meaning from the point of view that one benefit is greater than another .

Finally, there is no natural order for the results, you can put Better health care first instead of More money, for example, and it does not change the result at all.

Country of birth of a person

The country of birth is a nominal variable whose values ​​are the names of the countries. For purposes of working with this variable, it is convenient to make a numerical coding of this information, we assign code 1 to those born in Argentina, code 2 to Bolivia, code 3 to Canada, and so on.

This coding facilitates counting by computer and the management of data collection instruments. However, since we have assigned numbers to the various categories, we cannot manipulate these numbers. For example, 1 + 2 does not equal 3; that is, Argentina + Bolivia does not result in Canada.

References

Coronado, J. (2007). Measurement scales. Paradigms Magazine. Recovered from unitec.edu.co.
Freund, R.; Wilson, W.; Mohr, D. (2010). Statistical methods. Third ed. Academic Press-Elsevier Inc.
Glass, G.; Stanley, J. (1996). Statistical methods not applied to the social sciences. Prentice Hall Hispanoamericana SA
Beautiful.; Marchal, W.; Wathen, S. (2012). Statistics applied to business and economy. Fifteenth ed. McGraw-Hill/Interamericana Publishers SA
Orlandoni, G. (2010). Statistical measurement scales. Telos Magazine. Retrieved from ojs.urbe.edu.
Siegel, S.; Castellán, N. (1998). Non-parametric statistics applied to the behavioral sciences. fourth ed. Editorial Trillas SA
(2019). Level of measurement. Retrieved from en.wikipedia.org.

Deja una respuesta

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *