Multivalue

Multivalue refers to a property of data items that may contain more than one value. A multivalued attribute or field stores several elements within a single attribute location, in contrast to a single-valued attribute that holds exactly one value. Multivalue data appears in various computing contexts, including data models, databases, and data interchange formats.

In data modeling, multivalued attributes are seen in early entity-relationship theory and in some non-relational designs. In formal normalization, multivalued attributes are typically decomposed into separate relations to enforce atomicity, reduce redundancy, and simplify updates. For example, a person with multiple phone numbers is often represented by a separate phone numbers relation rather than storing all numbers in one field.

Some database systems, notably MultiValue databases such as Pick, UniVerse, and UniData, natively support multivalue fields. In these systems a field can hold a list of values or subfields, enabling direct storage of one-to-many relationships and specialized query and update semantics that differ from traditional relational databases. This design offers compact storage and particular tooling but can complicate cross-system interoperability and standardization.

In programming languages, multivalued data is implemented as arrays, lists, sets, or other collection types. Multivalued fields appear in forms and data formats where a single attribute may collect multiple inputs, requiring parsing and validation and, if used in relational contexts, careful normalization. Understanding whether to model data as multivalued depends on requirements for querying, updating, and maintaining data integrity.