Inputs
The inputs are the business case, the data and metadata, and the stakeholder section of the profile.
What clause 7.2 asks
Three inputs arrive: the business requirement, the data with its metadata, and the stakeholder part of the profile. The first supplies the setting and the intended population; the second supplies the material; the third supplies the people against whom representation is judged.
Why it matters
The inputs set the direction of the work. A team that starts with the data will produce a technically rich record that cannot say who is missing. A team that starts with the business case and the stakeholder set will ask which data is needed, which data exists, and which people are invisible in what exists. The metadata input is also a check on the data: if the metadata is thin, the provenance work in clause 7.4.2.1 will expose it.
What good looks like
Assemble the three inputs before the analysis. Read the business case for the context and the affected population. Read the stakeholder section for the groups and attributes. Then profile the data against both. Record any mismatch between the data the business case assumes and the data that exists, because that mismatch is often the first finding.
Evidence an assessor asks for
The assessor will ask which version of the stakeholder set was used. The answer should be a version, not a name.
- Business case extract The context of use and the intended population
- Data and metadata pack The data sets, their metadata and their known limitations
- Stakeholder extract The reference set and attributes against which representation is judged
- Gap note Any mismatch between the data the case assumes and the data that exists
Where it commonly fails
- The analysis starts from the data catalogue
- The business case is used only for the benefit figure
- The metadata is missing for the largest data set
What to do before publication
Bring the stakeholder extract and the business case to the first data session. They decide what the analysis is for.