“The quality and integrity of institutional data and institutional data products shall be actively managed.” – UW-Madison’s Institutional Data Policy (UW-523)
Dimensions of data quality
The standards and metrics of data quality are contextual and may differ across the various departments/units of an institution. The quality of data is determined based on the extent to which it is fit for purpose and may include (from DAMA’s Data Management Body of Knowledge):
- Accuracy: The degree that data correctly represents ‘real-life’ entities
- Completeness: Whether all required data is present
- Consistency: Ensuring that data values are consistently represented within a data set and between data sets, and consistently associated across data sets
- Integrity: Avoiding data loss, corrupted data, and orphaned data
- Reasonability: Whether a data pattern makes sense within its context
- Timeliness: Includes how frequently data values change, whether the data is up-to-date, and the time between when the data was created and when it became available for use
- Uniqueness/Deduplication: No entity exists more than once within the data set
- Validity: Whether data values are consistent with a defined domain of values
Benefits
- Critical to university functions and business processes
- Improves the accuracy and efficiency of the processes that rely on the data
- Enhancing trust and decision-making
- Reducing risks and losses associated with incorrect or unavailable data
Potential Risks
- Incorrect, obsolete, duplicate, or missing data
- Process does not operate as intended
- Changes occur in the data so that downstream reports or products that rely on the data are no longer accurate
Report a data quality issue
The university’s Institutional Data Issue Management Procedure provides a process for the intake, management, and resolution of institutional data issues.