A representative sample is a subset whose relevant characteristics adequately reflect the population from which it is drawn. In pre-project work, it provides a feasible basis for investigating data when obtaining the full dataset is technically, financially, contractually, or operationally impractical. section 2.4
Obtaining the sample is itself a discovery task: it can expose gaps in the customer’s understanding of the data, infrastructure, access constraints, and data-management effort. Greater early investigation reduces later delivery risk. section 2.4
Use a sample review checklist:
- Check coverage across the period in which the data was accumulated and across all source systems. section 2.4
- Inspect extreme values, value ranges, sparsity, repeated values, and the statistical properties of the sample. section 2.4
- Record how collection occurred, including surveys, experiments, business-process exhaust, regular intervals, or event-driven capture. section 2.4
