Assessment · 27
Data quality review
An assessment of whether the available data is suitable for its intended use.
When this is useful
Data can be adequate for one use and unsuitable for another. Quality checks need to relate defects to the analysis or process they affect.
My contribution
I examine the relevant records, assess the consequences of the problems found and recommend corrections. Any work to organise or correct the data is agreed separately.
Four presence checks on 10,000 fictional orders; 1,800 lack an account ID, affecting customer-level analysis.
What it covers
- Missing, duplicate or outdated records
- Conflicting definitions, values and sources
- Coverage and consistency checks
- Effects on the intended analysis or application
- Recommended corrections and further checks
The output
A review of data quality, its effects on the intended use and recommended preparation or correction work.
For an enquiry, describe the question and the systems involved.
contact@fiftystork.com ↗