Nominal vs Ordinal data: What's the difference for precise analysis?
Today, I’ll focus on the categorical data types, nominal vs ordinal data. I will also discuss their differences, uses, and everything in between.
According to finances online, a person on the Internet produces 1.7 MB of data per second. Every day, 2.5 quintillion bytes of data is generated online. To make sense of it all, categorizing data into different types is an absolute necessity.
Back in 1946, psychologist Stanley Stevens decided to organize data into neat little boxes. He labeled them nominal, ordinal, interval, and ratio, each defined by its level of measurement precision.
Nominal and ordinal data are both types of qualitative data, whereas interval and ratio data are classified as numerical/quantitative.
While each type offers distinct analytical possibilities, today, we’ll focus on the categorical data types, nominal vs ordinal data, their differences, uses, and everything in between.
What is nominal data?
Nominal data is a type of data used to label variables without providing any quantitative value. It falls under the category of qualitative data, which means it describes characteristics or qualities that cannot be measured numerically.
These variables are non-parametric, indicating that they don't rely on any assumptions about the distribution of the data.
Nominal data uses labels or names to identify different categories. These labels do not have any inherent order or ranking.
Each data point in nominal data belongs to one and only one category so that there is no overlap between the categories.
Nominal data does not have any numerical value or order. You cannot perform mathematical operations like addition or subtraction on nominal data. For example, you can't say that blue is more than red.
Since nominal data is qualitative, it focuses on describing attributes or qualities rather than measuring them.
Businesses use nominal data to categorize customer preferences. Customer’s favorite brands is a nominal data example. It is also commonly used in surveys to collect demographic information such as gender, nationality, or marital status.
Nominal data example: Employee department assignment
In this nominal data example, I categorized employees by their department within a company. These departments are labels without any inherent order or ranking.
Example:
- Marketing
- Sales
- Human Resources
- IT
What is ordinal data?
Ordinal data is used to represent categorical variables that have a natural order or ranking. It is also qualitative and non-parametric, which means you can't measure it in numbers, and it doesn't assume a specific distribution for the data.
The primary feature of ordinal data is that it has a clear, natural order. The categories can be ranked or arranged in a specific sequence. A customer satisfaction survey with answer options like "very dissatisfied," "dissatisfied," "neutral," "satisfied," and "very satisfied” measures ordinal data.
While the data points are ordered, the intervals between them are not necessarily equal. This means you can say one category is higher or lower than another, but you can't quantify the exact difference.
For example, the gap between "very satisfied" and "satisfied" may not be the same as the gap between "neutral" and "dissatisfied.”
Ordinal data is commonly used in surveys to gauge opinions, preferences, or levels of agreement. In educational settings, you can use it to rank students based on grades or performance levels.
Ordinal data example: Employee performance rating
In this ordinal data example, I ranked employee performance levels. These ratings have a natural order, but the intervals between them are not necessarily equal.
Example:
- Poor
- Fair
- Good
- Very Good
- Excellent
These were the definitions of ordinal and nominal data. Let’s look at their differences next.
Nominal vs Ordinal data: 6 key differences
Now, let’s look at what sets ordinal vs nominal data apart in detail.
Set order
When comparing nominal and ordinal data, one key difference is the presence of a set order.
Nominal data does not have any set order. The labels used in nominal and categorical data are simply names without any ranking or sequence.
If you categorize people by their favorite ice cream flavors—chocolate, vanilla, and strawberry—there is no inherent order to these flavors. One flavor is not considered higher or lower than another. The categories are equal and independent of each other.
Ordinal data, on the other hand, has a clear order or ranking. It organizes categories in a specific sequence, where one category is higher or lower than another.
For instance, if you rate your satisfaction with a service as poor, fair, good, or excellent, these ratings have a clear order. "Poor" is less than "Fair," which is less than "Good," and "Excellent" is the highest rating.
Quantitative value
Nominal data does not have any quantitative value. It only labels categories without assigning any numerical value to them. You cannot perform mathematical operations like calculating the mean or standard deviation on nominal data because there are no numerical values.
With nominal data, you can find the margin of error and the mode, which is the category that appears most often. However, you cannot calculate the mean or median because these require numerical values.
Ordinal data sits between categorical and quantitative data. It provides a ranking or order, but the intervals between the ranks are not necessarily equal.
Because ordinal data has a ranking, you can calculate the median, which is the middle value when the data is ordered. If you have an odd number of ratings, the median is the middle one. If you have an even number, it’s the average of the two middle values.
You can also find the mode, just like with nominal data. However, calculating the mean or standard deviation is not appropriate because ordinal data does not have equal intervals.
Tests you can perform
Different types of statistical tests are performed on nominal vs ordinal data.
For nominal data, the focus is on frequency and proportion. Here are the tests applicable:
- McNemar test checks if paired data changes over time.
- Cochran’s Q Test compares multiple paired groups.
- Fisher’s exact test determines the relationships between two categorical variables in small datasets.
- Chi-Square test examines the relationship between two categorical variables
Here are four key statistical tests you can use with ordinal data:
- Wilcoxon signed-rank test compares two related groups.
- Friedman 2-way ANOVA compares multiple related groups.
- Wilcoxon rank-sum test compares two independent groups.
- Kruskal-Wallis 1-way test compares multiple independent groups.
Data collection methods
When collecting data, the methods you use can differ based on whether you are gathering nominal vs ordinal data.
For nominal data, you can use drop-down, checklists, or multiple choice questions to ask people about their preferences, characteristics, or behaviors. During interviews, you can also ask open-ended questions that classify responses into categories.
Since ordinal data involves categories with a specific order, Likert/rating scales or rating questions are used in data collection surveys. In educational or workplace settings, you can use performance assessments to rank individuals based on their performance levels.
Descriptive analytics
Descriptive analytics helps summarize and visualize nominal vs ordinal data and turn them into actionable insights.
For nominal data, which involves categories without any specific order, bar charts, and pie charts are commonly used.
Bar charts display the frequency of different categories, while pie charts show the proportion of each category within a whole. Frequency tables are also useful for listing each category along with the number of times it appears.
For ordinal data, histograms and box plots are the most effective. Histograms show the frequency of ordered categories, and box plots provide a visual summary of the distribution, including the median and quartiles.
Predictive analytics
Predictive analytics uses past data to predict future events. For nominal data, which consists of categories without a specific order, decision trees, logistic regression, and association rules are commonly used techniques. However, predictive analysis is very limited because there are no numbers to play with.
Ordinal logistic regression predicts the likelihood of outcomes in ordinal data based on ordered categories, such as predicting a student's grade based on their previous performance. Survival analysis predicts the time until an event occurs, like customer churn or equipment failure.
Use cases of Nominal Vs. Ordinal data
Now, let's talk about where you actually see nominal vs. ordinal data in the real world.
There are countless advantages of nominal data. Some areas where nominal data is useful include:
- Market Research: Companies often use nominal data to understand customer preferences.
- Healthcare: In the medical field, nominal data is used to categorize patients by blood type, gender, or the presence of certain conditions.
- Elections: During elections, nominal data is used to record voter preferences for different candidates or parties.
Ordinal data, in contrast, is used for:
- NPS and CSAT surveys: Ordinal data is used in surveys where respondents rank their satisfaction on a scale.
- Education: In schools, ordinal data is used to grade students' performance, such as letter grades (A, B, C, etc.) or ranking students in a class. It is also used in student perception surveys.
- Job Performance: Companies use ordinal data to evaluate employee performance with ratings like "poor," "fair," "good," and "excellent."
Nominal data examples
Gender, country of origin, and blood type are a few examples of nominal variables. These categories are named but do not have numeric values associated with them.
Here are some more nominal data examples:
- Colors: Red, blue, green, yellow
- Genders: Male, female, non-binary
- Countries: United States, Canada, Mexico, Australia
- Car types: Sedan, SUV, truck, sports car
- Marital status: Single, married, divorced
- Eye color: Brown, blue, green, hazel
- Hair color: Black, brown, blonde, red
- Pet types: Dog, cat, bird, fish
- Occupation: Teacher, doctor, lawyer, engineer
Example of ordinal data
Ordinal data similarly has categorical values, but these values are arranged in a logical ranking or progression. Anything that can be listed according to an order is an example of ordinal data.
Some of these include:
- Education level: High school, college, master's, doctorate
- Job title: Entry-level, junior, senior, manager
- Product rating: 1 star, 2 stars, 3 stars, 4 stars, 5 stars
- Likert scale responses: Strongly disagree, disagree, neutral, agree, strongly agree
- Economic status: Low income, middle income, high income
- Product size: Small, medium, large, extra large
- Age groups: Child, teenager, adult, senior
Formaloo, the best tool to collect nominal and ordinal data!
Both nominal vs ordinal data fall under the categorical umbrella. Their unique characteristics influence how they're collected, analyzed, and interpreted. Effectively using the right statistical tests and visualization techniques for each data type is essential for drawing meaningful conclusions.
Formaloo, a customizable survey maker, empowers you to effortlessly gather nominal and ordinal data through customizable questionnaires.
Beyond data collection, Formaloo offers robust survey analytics capabilities to help you visualize and interpret your findings with features like data filtering, cross-tabulation, and trend analysis.
Try Formaloo for free today!