The selected attributes are measured with the designated measurement method, manually (a defect form sent to the analyst) or automatically (a requirements tool logging each update). Collected data is stored together with the context information needed to verify, understand or evaluate it, in a store that may be paper-based for a small short-lived effort and is often managed through the information management process. Data is verified, usually by inspection against a checklist that confirms missing data is minimal and values make sense (a valid defect class, a component size not ten times every previous one) and by automated range and type checks; anomalies are taken up with the data providers and raw data corrected where necessary. The analyst carries responsibility for verification, working with the providers.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.