Skip to main content

Attribute types: nominal, ordinal, interval, ratio

Short answer

An attribute is a property of a data object (e.g., age, color). Four types: Nominal (labels only, equality), Ordinal (ordered labels, ranking), Interval (ordered, meaningful differences, no true zero), Ratio (true zero, meaningful ratios). Nominal/ordinal are categorical; interval/ratio are numeric. Each type permits operations of weaker types.

What do you mean by attribute? Explain Nominal, Ordinal, Interval and Ratio attribute types.

An attribute (also called variable, feature or field) is a property or characteristic of a data object, for example the age, colour or marks of a student. The set of attribute values describes one object (record).

The four classical types, from weakest to strongest, differ in which operations are meaningful.

TypeMeaningOperationsExamples
NominalNames/labels only; no order=, ≠ (mode)eye colour, gender, zip code, ID number
OrdinalLabels with a meaningful order; gaps not meaningful=, ≠, <, > (median)grade (A, B, C), rank, small/medium/large
IntervalOrdered, differences meaningful, no true zero+, − (mean, std. dev.)temperature in °C or °F, calendar dates
RatioInterval plus a true zero, so ratios meaningful+, −, ×, ÷height, weight, age, salary, length, counts

Points to remember

  • Nominal and ordinal are categorical (qualitative); interval and ratio are numeric (quantitative).
  • 40 °C is not "twice as hot" as 20 °C because 0 °C is arbitrary (interval); 40 kg is twice 20 kg because 0 kg means no weight (ratio).
  • Each type allows all the operations of the types above it in the table, so a ratio attribute can be treated as interval, ordinal or nominal, but not the reverse.
  • The type decides which statistics and which mining algorithms can be applied.

✓ Answer checked by Claude Opus 5.5 (high effort)

Related questions

All Data Mining chapter 2 questions →