Approaches to Deriving Data Quality Definitions

Overview

Data quality definitions emerge from three distinct methodological traditions:

Approach	Method	Strengths	Weaknesses
Intuitive	Expert judgment, practitioner experience	Practical, fast to develop	Lacks rigor, hard to validate
Theoretical	Ontological analysis, information theory	Rigorous foundations	May miss practical concerns
Empirical	User studies, surveys of data consumers	Grounded in actual needs	Context-specific, may not generalize

When Each Applies

Intuitive approaches work when:

• Quick definition is needed for a specific project
• Deep domain expertise is available
• The context is well-understood and stable

Theoretical approaches work when:

• Foundational definitions are needed that can be built upon
• The goal is cross-domain applicability
• Formal reasoning about quality is required

Empirical approaches work when:

• User-facing quality matters most
• The definition must reflect actual (not assumed) quality needs
• Validation against real-world usage is critical

Key Differences

Foundation: Intuitive definitions rest on experience; theoretical on formal models; empirical on data from users.

Validation: Intuitive definitions are hard to validate objectively; theoretical definitions can be checked for internal consistency; empirical definitions can be validated against user satisfaction.

Generalizability: Theoretical definitions aim for universality; empirical definitions reflect the population studied; intuitive definitions reflect the expert’s experience.

The Pattern

The most robust quality frameworks combine approaches: theoretical foundations establish the space of possible dimensions, empirical studies identify which dimensions matter to actual users, and practitioner intuition fills gaps and guides prioritization.

Related: 04-atom—data-quality-consensus-gap, 03-molecule—foda-taxonomy-methodology, 04-atom—fitness-for-use-definition