Types of Attributes: Nominal, Ordinal, Interval and Ratio
An attribute is a property or characteristic of an object. The four attribute types differ by which properties they support: distinctness, order, meaningful differences and meaningful ratios.
Each type adds a property to the one before it. Nominal attributes support only distinctness (= and not equal): the values are just names or labels, such as eye colour, ID numbers and zip codes. Ordinal attributes add order (<, >): rankings, grades, shirt sizes. Interval attributes add meaningful differences (+ and -): calendar dates and temperature in Celsius or Fahrenheit, where there is no true zero. Ratio attributes add meaningful ratios (* and /): temperature in Kelvin, length, age, counts and money, where zero means none.
Nominal and ordinal attributes are categorical or qualitative; interval and ratio attributes are numeric or quantitative. Another split is by number of values: a discrete attribute has a finite or countably infinite set of values, which may or may not be integers (zip codes, counts), while a continuous attribute takes real numbers (temperature, height). Binary attributes are a special case of discrete attributes. In an asymmetric attribute only the presence (a non-zero value) matters, for example which words appear in a document.
A dataset of records can be represented by an m by n matrix, with m objects (rows) and n attributes (columns). Graphs, molecules and web pages are examples of graph-based data.
Key points
- Nominal: distinctness. Ordinal: plus order. Interval: plus differences. Ratio: plus ratios.
- Interval has no true zero; ratio does.
- Nominal and ordinal are categorical; interval and ratio are numeric.
- Discrete: finite or countably infinite values. Continuous: real numbers.
- Asymmetric attribute: only non-zero (presence) values matter.
Worked example
Is 40 C twice as hot as 20 C? No: Celsius is interval, so the ratio is not meaningful. 400 K is twice 200 K because Kelvin is a ratio scale. A star rating 1 to 5 is ordinal because order is meaningful but the gap between 1 and 2 is not necessarily equal to the gap between 4 and 5.
Common questions
What is an attribute?
A property or characteristic of an object, also called variable, field, feature or dimension.
Why is temperature in Celsius interval, not ratio?
Zero degrees Celsius does not mean no temperature, so ratios such as twice as hot have no meaning.
Practice questions on this topic
Try these, then open the app to check your answers with explanations.
- Examples of interval attributes can be:
- A. Calendar dates
- B. colors
- C. ID Numbers
- Data that consists of a collection of records, each of which consists of a fixed set of attributes
- A. Graph Data
- B. Ordered Data
- C. Record Data
- The following activity is a data mining Task: Predicting the future stock price of a company using historical records.
- A. false
- B. true
- What are the different types of attributes?
- A. Ordinal
- B. Interval
- C. Nominal
- D. All the above
- Examples of Ordinal attributes can be:
- A. ID Numbers
- B. Satisfaction rating (poor / fair / good / excellent)
- C. zip codes
- Suppose that the data for analysis includes the attribute grade. The grade values for the data tuples are: 4, 5, 9, 11, 12, 13, 13, 13, 14, 15, 16, 17, 18, 18, 19, 20. What is the median of the data?
- A. 13.5
- B. 13.6
- C. 13.7
- D. 13
Related topics
- What is Data Mining? Tasks: Classification, Clustering, Regression, Association Rules
- The KDD Process: Knowledge Discovery in Databases Steps
- Data Cleaning: Noise, Outliers, Missing Values and Inconsistent Data
- Binning: Equal-Width vs Equal-Frequency (Equal-Depth) with Examples
- Data Normalization: Min-Max, Z-Score and Decimal Scaling with Examples
- Quartiles, IQR and the 1.5 x IQR Outlier Rule
- All topics: Data Mining
BASMAGLY is an independent student project, not an official Delta University site. Check lectures and your instructor for the final word.