Ordinal variables are variables that have categories with a specific order or ranking to them, but the distances between the categories are not known or consistent (Babbie et al., 2007).
Examples include rating scales like “low”, “medium”, and “high” or education levels such as “elementary”, “high school”, and “college”.
The key feature of ordinal variables is that they provide specific information about the position or order of the variables (Katz, 2006b; Stockemer, 2018). However, they generally don’t demonstrate distances between intervals on the scale, meaning researchers need to be discerning about whether ordinal variables are best for their research project.
Ordinal Variables Examples
1. Education Level
Education level is often considered an ordinal variable. Typical categories might include “elementary”, “middle”, “high school”, “college”, and “post-graduate”. The categories imply a certain order or ranking in terms of academic achievement, but the distance between categories (e.g. the difference in learning between “high school” and “college”) is not uniformly measured or consistent.
2. Socioeconomic Status
Socioeconomic status is another example of an ordinal variable. This could be categorized into “low income”, “middle income”, and “high income”. Here, the categories are in a specific order but the differences between categories aren’t well-defined or consistent. The gap of income between “low” and “middle” income might be significantly different than that between “middle” and “high” income.
3. Likert Scale
Often used in surveys, a Likert scale is an ordinal variable that allows respondents to indicate their level of agreement or satisfaction, typically on a 1 to 5 or 1 to 7 scale. “Strongly disagree”, “Disagree”, “Neutral”, “Agree”, and “Strongly Agree” might be the categories. The specific numerical distance between each response is not known.
4. Movie Ratings
A movie rating is a common ordinal variable. A film could be rated as “poor”, “average”, “good”, and “excellent”. While these categories provide an order, the difference between a “poor” and “average” movie isn’t necessarily of the same magnitude as the difference between an “average” and “good” movie.
5. Military Ranks
Military ranks, such as “private”, “corporal”, “sergeant”, “lieutenant”, and “colonel”, represent an order or ranking. However, the extent of the difference in authority or responsibility between each rank is not uniformly measured or consistent.
6. Pain Scale
In medical contexts, a pain scale is often used as an ordinal variable. Patients are asked to rate their pain on a scale of 1 to 10, with 10 representing the most severe pain. This provides a rank order of pain, but the difference in perceived pain between each number isn’t uniformly defined.
7. Restaurant Ratings
Similarly, restaurant ratings are ordinal variables. A restaurant might be given one to five stars based on food quality, service, and ambiance. The difference in quality between each star rating isn’t uniformly defined or consistent.
8. Customer Satisfaction
Customer satisfaction is typically measured as an ordinal variable. Customers might be asked if they are “dissatisfied”, “neutral”, or “satisfied”. These categories offer an ordered way to evaluate satisfaction, but the difference in feeling between each category is not uniformly measured.
9. Product Quality
The quality of a product can also be rated as an ordinal variable. Consumers might rank a product quality as “poor”, “average”, “good”, or “excellent”. These categories offer an ordered way to perceive quality, but the difference between each is not consistently defined.
10. Risk Level
Lastly, scenarios that involve assessing risks often employ ordinal variables. “Low risk”, “medium risk”, and “high risk” could represent categories. The categories convey a precise order of risk, but the specific degree of difference or magnitude between categories isn’t known or uniform.
11. Clothing Sizes
Clothing sizes typically are described by the ordinal variables of “Small”, “Medium”, “Large”, and “Extra Large”. These categories represent an inherent sequence but the physical differences between them vary from brand to brand.
12. Book Ratings
Any rating system where a book is rated on a scale for its content is ordinal. For example, a one to five star evaluation scale. While the scale suggests order — five stars is better than four — the difference between a four and a five-star rating would be subjective and not uniformly defined.
13. Job Satisfaction
Often surveyed on a scale of “very dissatisfied”, “somewhat dissatisfied”, “neutral”, “somewhat satisfied”, “very satisfied.” These categories provide a clear order but the difference between them isn’t well-defined or quantifiable.
14. Confidence Levels
When asked about confidence in performing a task, respondents could signify their answer as “not confident”, “somewhat confident”, “confident”, or “very confident”. There is a clear order to the categories, but no specified numerical distance between them.
15. Frequency of Occurrence
It is common to use terms like “Never”, “Rarely”, “Sometimes”, “Often”, and “Always” to denote the frequency of occurrence, making it an ordinal variable. Although these categories suggest a ranking, the exact time duration they represent isn’t uniformly defined.
16. Difficulty Level
The difficulty level of tasks or exercises, commonly categorized as “Easy”, “Intermediate”, “Hard”, and “Extreme”, are examples of ordinal variables. The categories indicate an order but the difference in complexity or effort between them isn’t well-defined or evenly spaced.
17. Hotel Class
The star rating for hotels (one-star, two-star, three-star, etc.) is an ordinal variable. The star ranking provides a categorical indication of quality, but the actual quality difference between each star level is not uniformly measured or defined.
18. Severity of Symptoms
In a medical survey, patients could be asked to rate the severity of their symptoms as “Mild”, “Moderate”, or “Severe”. These categories have an inherent order, but the exact degree of discomfort each category represents isn’t uniformly defined.
19. Leadership Ratings
When rating leadership skills, categories such as “excellent”, “average”, and “poor” can be used. These states provide a clear sequence but the specific characteristics qualifying for each category can vary.
20. Company Ranking
Companies of the same industry may be ranked as “Top”, “Middle”, and “Low” performers, making this an ordinal variable. There is a clear order, but the performance distance between the categories is not uniformly measured.
21. Course Difficulty
Classroom courses are often subjectively described as “easy”, “medium”, or “difficult”. This ranking order does not provide consistent, measureable differences between each label.
22. Employee Performance
Employee performance grading is typically ordinal. Categories can include “unsatisfactory”, “meets expectations”, “exceeds expectations”. There is a clear order, but the exact amount of extra work done to exceed expectations is not quantifiable.
23. Age Group
A division of individuals into “young”, “middle age”, and “old” represents an ordinal variable where actual age difference between categories is not consistent.
24. Team Rankings
In sports, teams are often categorized into “first division”, “second division”, etc. These categories represent an order, but the difference in skill or performance between each isn’t quantified or uniformly defined.
25. Social Class
Divisions of “lower”, “middle”, and “upper” class represent an ordinal variable. These categories indicate an order of social class, but the wealth difference between each class is not consistently defined.
Types of Variables (Compare and Contrast)
Ordinal variables are generally contrasted to nominal, interval, and ratio variables as the main types of variables.
Each is explained below.
- Nominal variables represent categories that don’t have a natural order or ranking (Wilson & Joye, 2016). Brands of cereal or types of music are examples of nominal variables. One must remember that you cannot perform mathematical operations on nominal variables. The only thing you can do is count how many instances fall into a specific category.
- Interval variables are variables having ordered categories with a known and consistent distance between them (Lewis-Beck, Bryman & Liao, 2004). Temperature measured in degrees Celsius, where the difference between 30 and 31 is the same as between 20 and 21, serves as a classic example of interval variables. However, because there’s no true “zero point”, it’s not advisable to make comparisons of magnitude with interval variables. You can’t say that 30°C is ‘twice as hot’ as 15°C.
- Ratio variables are akin to interval variables, but they possess a clear definition of zero (Katz, 2006a; Katz, 2006b). Examples of ratio variables are height in centimeters or weight in pounds. Here, zero implies the total absence of a quantity, allowing one to make definitive statements like “Person A weighs twice as much as Person B”.
- Ordinal variables are a type of categorical data where the categories carry a sense of order or ranking (De Vaus, 2001). This integral component distinguishes them from nominal variables. Think of a race where athletes finish in “first”, “second”, “third” positions (these are their ranks). However, the exact numerical distances or intervals between ordinal categories remain unknown or inconsistent, unlike in interval or ratio variables.
Ordinal variables play an indispensable role in quantitative research across a wide array of disciplines. These variables enable the categorization of data in an ordered format, serving as a bridge between nominal variables (having categories without an order) and interval/ratio variables (having categories with both an order and known distances). However, while ordinal variables provide a sense of order, the inability to ascertain the exact or consistent distance between different categories presents a limitation. The implications of ‘jumping’ from one category to another aren’t always definitively measurable or comparable, given the absence of uniform spacing between categories. As such, while the sequential nature of ordinal variables offers distinct advantages in collating and interpreting data, the lack of numeric specificity restricts the kind of statistical analysis that can be performed. Given these benefits and drawbacks, the choice to use ordinal variables in research should align with the nature of data being studied and the research goals at hand.
Babbie, E., Halley, F., & Zaino, J. (2007). Adventures in Social Research: Data Analysis Using SPSS 14.0 and 15.0 for Windows (6th ed.). New York: SAGE Publications.
De Vaus, D. A. (2001). Research Design in Social Research. New York: SAGE Publications.
Katz, M. (2006). Study Design and Statistical Analysis: A Practical Guide for Clinicians. Cambridge: Cambridge University Press.
Katz, M. H. (2006). Multivariable analysis: A practical guide for clinicians. Cambridge: Cambridge University Press.
Lewis-Beck, M., Bryman, A. E., & Liao, T. F. (Eds.). (2004). The SAGE Encyclopedia of Social Science Research Methods (Vol. 1). London: SAGE Publications.
Stockemer, D. (2018). Quantitative Methods for the Social Sciences: A Practical Introduction with Examples in SPSS and Stata. London: Springer International Publishing.
Wilson, J. H., & Joye, S. W. (2016). Research Methods and Statistics: An Integrated Approach. New York: SAGE Publications.
Dr. Chris Drew is the founder of the Helpful Professor. He holds a PhD in education and has published over 20 articles in scholarly journals. He is the former editor of the Journal of Learning Development in Higher Education. [Image Descriptor: Photo of Chris]