You've read that data quality has 6 dimensions. You've also seen frameworks with 10, 12, even 15 dimensions. They all claim to be comprehensive. Which one should you actually use?
The short answer: the one your team will consistently apply. The dimensions framework is a diagnostic tool, not a compliance checklist — and a simpler framework used consistently beats a comprehensive one used never.
DAMA's 6 Core Dimensions
The Data Management Association (DAMA) is the most widely cited authority on data quality. Their framework identifies six core dimensions:
- Accuracy — Does the data reflect reality?
- Completeness — Are all required values present?
- Consistency — Is the same fact represented the same way everywhere?
- Timeliness — Is the data current enough for its use case?
- Validity — Does the data conform to defined formats and business rules?
- Uniqueness — Does each entity appear exactly once?
Sohovi profiles every column in your dataset for completeness and flags the exact rows where values are missing — free to try.
These six cover the vast majority of data quality problems encountered in practice. For most small and mid-size organizations, this is the right framework.
Extended 10+ Dimension Frameworks
Some frameworks add additional dimensions:
- Integrity — Relationships between records are correct (referenced records exist)
- Conformity — Values follow defined format standards
- Precision — Data carries the appropriate level of detail
- Relevance — Data is appropriate for the current use case
- Accessibility — Data is available when needed
These additional dimensions are valuable for specific use cases. Integrity matters for relational database management. Precision matters for scientific and financial data. Relevance matters for AI training datasets.
Sohovi profiles your training data for quality issues — missing values, outliers, type mismatches — before they corrupt your model.
Which Framework to Use
Use the 6-dimension DAMA framework if:
- You're new to data quality and need a practical starting point
- Your team is non-technical and needs a simple shared vocabulary
- You're doing a general-purpose data audit
Use an extended framework if:
- You're managing complex relational data with many system integrations
- You're building data quality into a technical data pipeline
- You have specific requirements (financial precision, regulatory compliance) that the core six don't fully address
A Practical Decision Rule
Start with the 6 DAMA dimensions. Score your most important datasets against each. If you identify problems that the 6 dimensions don't fully capture, add the specific additional dimensions that address those gaps.
Tools like Sohovi profile datasets against the core dimensions automatically — completeness, uniqueness, validity, and consistency — giving you a starting score without choosing a framework first.
The goal isn't framework compliance. It's data that's fit for purpose. Use whatever framework helps you get there.
