DataTerminology.comNEW
Also known as: Ordered categorical data
Categorical data with an inherent order but unknown or unequal intervals between categories.
DESCRIPTION
Ordinal data represent categories with a meaningful order, as in ratings from poor to excellent. Distances between categories are undefined, so medians and rank-based tests are appropriate. Common examples include Likert scales and socioeconomic classes. Modeling often uses ordinal logistic or probit regression. Encoding choices affect downstream models; avoid treating ordinal values as interval without justification.
ALTERNATIVE DEFINITIONS
Wikipedia
Ordinal data is a categorical, statistical data type where the variables have natural, ordered categories and the distances between the categories are not known.
Highlights ordered categories with unknown intervals.
Penn State (STAT 200)
Ordinal variables have categories with a meaningful order, but the differences between categories are not uniformly meaningful.
Explains interpretability of order versus differences.
NIST/SEMATECH e-Handbook
Ordinal data: data that can be categorized and ranked, but the distances between ranks cannot be assumed equal.
Stresses rankability without equal-interval assumption.
RELATED TERMS
Ratio Data Likert Scale Ordinal Regression Nonparametric Statistics Ranking Categorical Data Measurement Scale Median
COMPETING TERMS
Nominal Data Interval Data
RELATIONS
- Broader: Data
KNOWLEDGE AREAS
Data Science Data Management Data Modeling
TARGET AUDIENCE
Data Scientist Data Analyst Data Steward Information Analyst
EXTERNAL SOURCES
- Wikipedia - Ordinal data
- NIST - Measurement scales overview
- Penn State - Types of data and scales
- UCLA Statistical Consulting Group - Data types guidance
TOP QUESTIONS
What statistics are appropriate for Ordinal Data?
Use medians, percentiles, and nonparametric tests (e.g., Mann–Whitney). Means and standard deviations can mislead.
Category: Best practices
How should Ordinal Data be encoded for ML?
Use ordinal encoders preserving order or target encoding with care; avoid one-hot if order is essential.
Category: Requirements
What is a common mistake with Ordinal Data?
Treating ordinal categories as interval-scaled and applying inappropriate arithmetic operations.
Category: Mistakes
What models fit Ordinal Data outcomes?
Ordinal logistic or probit regression are common choices for ordered outcomes.
Category: Overview
Is Likert-scale data always ordinal?
Typically yes, though practitioners sometimes approximate it as interval under certain conditions.
Category: Misunderstanding
Why is Ordinal Data important?
Many surveys, ratings, and rankings produce ordered categories central to analysis and decision-making.
Category: Relevancy
How many categories are needed for Ordinal Data?
At least two ordered categories, though more levels often improve modeling granularity.
Category: Requirements
How to visualize Ordinal Data?
Use ordered bar charts, stacked bars, or cumulative distributions to respect order.
Category: Best practices
What is an example of Ordinal Data?
Customer satisfaction ratings from 1 (very dissatisfied) to 5 (very satisfied).
Category: Example