What Is Ordinal Data and How to Use It? (With Examples)

By Indeed Editorial Team

Published 24 October 2022

The Indeed Editorial Team comprises a diverse and talented team of writers, researchers and subject matter experts equipped with Indeed's data and insights to deliver useful tips to help guide your career journey.

Data collection is an important element of research. Ordinal data is one type of statistical measurement that researchers can collect through surveys to discover habits or feelings of respondents. Understanding ordinal data and its usefulness can be valuable if you're considering a research career. In this article, we answer the question 'What is ordinal data?', identify some reasons for using it, highlight the major categories of ordinal data, outline the key differences compared to other measurement types, provide some examples of its use and list some tools for collecting this type of data.

What is ordinal data?

When answering the question 'What is ordinal data?', it's important to understand that ordinal data is a type of categorical data that assigns specific variables to categories and ranks. This can mean there's no defined space or there's an unequal space between each category. The ranking is typically hierarchical, ranging from the highest to the lowest or the most frequent to the least frequent. Ordinal data is useful when undertaking qualitative research. Although the data itself isn't numerical, it may use numbers to group similar variables, which can help make it easier for researchers to understand and analyse their findings.

Related: 18 Qualitative Data Examples: With Definition and Types

Why use ordinal data?

Ordinal data scales can provide specific findings that can be helpful for business strategy. You typically see ordinal data in questionnaires or surveys, where researchers encourage respondents to answer using a range of answers. The following are a few reasons to use ordinal data when undertaking research:

Collecting demographic data

Demographic data can relate to participants selecting which group they belong to, which can help researchers track the number of participants for each group and look for patterns in their responses. Because demographic data can include a specific ranking system, it's a type of ordinal data. Some examples can comprise education, income or age of respondents.

Related: What Is a Target Demographic? (Plus How to Identify Yours)

Satisfaction ratings

If you're interested in understanding how your target audience feels about the organisation you work for, its product or its service, using satisfaction surveys to get ordinal data can be helpful. When respondents' choices provide a range, you can order these according to their values. Various industries may use questionnaires using ordinal data to help gauge satisfaction.

For example, call centres can ask for feedback from customers after a representative assists them with an enquiry or a retailer may seek online feedback from customers after completing a purchase. Their responses may feature on a scale from very satisfied to very dissatisfied or encourage responses such as strongly agree to strongly disagree to help organisations better understand what areas of their business require improvements.

Related: Customer Satisfaction Surveys (Including Example Questions)

Gauging levels of proficiency

Another way researchers can use ordinal data in questionnaires is to help assess a respondent's capability to handle a specific task. For example, an educator may ask students to self-assess their learning and development. Alternatively, a human resource (HR) manager may ask potential candidates how competent they are at specific skills or tasks.

Questions can also include different scale levels to help understand a respondent's level of experience. For example, a hiring manager may ask a potential candidate how many years they have worked in a specific industry on a job application. Set answers may include less than a year, one to three years or more than three years, depending on the manager's requirements.

Related: What's Industry Experience and When Do You Require It?

Different categories of ordinal data

Ordinal data typically includes two main categories:

Matched category

In a matched category, researchers group data samples with similar characteristics according to variables. Grouping is useful for estimating the differences in the data set. Eliminating variables helps to prevent influencing results. For example, matching age ranges or genders when investigating health issues typically more common in females or older people can be helpful.

Unmatched category

You can refer to these as unmatched samples or independent samples. Researchers can randomly select these samples using variables that don't depend on other ordinal variables. Researchers then base their analysis with the assumption that the samples are independent.

Related: A Guide to Data Analysis (With a Definition and FAQs)

Ordinal data vs. other types of data

In addition to ordinal data, researchers can use other types of statistical measures. The following are the main differences between the measurement types:

Nominal vs. ordinal

Nominal data organises different information into specific categories. In contrast to ordinal data, there's no definite way to rank nominal variables. For example, eye colour is a nominal variable. Researchers can't rank different eye colours in ascending or descending order. They're just different. Unlike nominal data, ordinal data can categorise various elements into sequences.

Ratio vs. ordinal

Like ordinal data, ratio data allows researchers to classify and order the various elements. There's also equal spacing between the components. Unlike ordinal data, ratio data elements can't be negative and there's no definite value of zero. Ratio data may be helpful to rank test results, where an educator could order scores from the highest to the lowest. Ratio data can also be beneficial in undertaking quantitative studies. Examples of ratio data can include measurements of heights, weights, sales or crime and unemployment rates.

Related: Skills Test (Definition and Examples)

Interval vs. ordinal

Interval and ordinal variables are somewhat similar statistical measures. Both types assign specific categories to data and rank data from the highest to the lowest. Like ordinal data, interval data can have negative values. The difference with interval data is that each rank has an identical value. Examples of interval data include measuring temperature, assessing IQs or probability studies.

Related: Types of Sampling (Including Definitions and Methods)

Examples of using ordinal data

The following are some specific examples of how ordinal data can be helpful for researchers:

Age group

Age demographics can be valuable to signify ordinal relationships. A researcher may split a group of respondents according to age. For example, younger participant categories may include children, teenagers or young adults, and older participants can be middle-aged or seniors. Age groups can represent ordinal data, which can be helpful if there's a significant age gap between categories that may influence the findings.

Socioeconomic status

Income levels can be a form of ordinal data. A researcher may ask respondents to include their socioeconomic status on a survey that organises levels from the lowest to the highest amount of income. Categories may comprise the lower, middle or upper class, with lower and middle classes having a smaller gap in annual income than the middle to the upper class. Although the categories are next to each other, the value between each type isn't the same.

Education

Education can be a typical example of ordinal data, allowing respondents to assign a specific rank against their education level. For example, a high school certificate may have the lowest rank and a doctorate may have the highest. Although the education level does increase from a high school certificate to a doctorate, the amount between each category isn't always the same. For example, there may be a more significant space between a high school certificate and a bachelor's degree than between an advanced diploma and a bachelor's degree.

Agreement

Surveys can often use scales to get respondents to indicate their position along a range. For example, a multiple-choice question may ask whether a respondent strongly agrees, agrees, is neutral, disagrees or strongly disagrees about a specific topic or view. The level of emotion between the ranges of the scale can be different. For example, the sentiment between strongly agree and agree is likely smaller than between agree and neutral. Because all values don't match, the answers on the scale are ordinal data.

Frequency

Researchers can format scale questions so respondents identify how often they undertake an activity. For example, a researcher may wish to understand teenagers' social media habits. So, they might survey teenagers to determine their social media platforms and how often they use them. Answer choices may range from the most frequent to the least frequent and respondents may choose categories such as very often or often to not often or never. Because there's no assigned value between each answer, the information collected is ordinal.

Language

Differing levels of language proficiency are another example of ordinal data. For example, when you start to learn a new language, you may classify yourself as a beginner. You may progress to an intermediate stage and become fluent as you practise. Because there's an uneven value between categories, the data is ordinal.

Ordinal data tools

Researchers can use specific tools to gather ordinal data, including the following:

  • Likert scale: Researchers use a point scale to seek feedback on someone's opinion about a particular subject. It's typically a five- or seven-point scale where options range from one extreme to the other such as very satisfied to very dissatisfied.

  • Interval scale: In this ordinal scale, each response is an interval. For example, classifying people into different groups according to age, such as youth or senior.

  • Slider scale: Also known as a continuous rating scale, a slider scale typically displays a horizontal slider beneath a question. Sliders allow a respondent to move the slider from left to ride and set a numerical value for their answer.


Explore more articles