A point materially outside established control limits is a classic signal of special-cause variation. Statistical Process Control distinguishes routine variation inherent in a stable process from unusual variation introduced by an identifiable event or change.
When a Data Quality process normally produces a stable percentage of invalid records and a sudden observation lies beyond the expected statistical range, investigation should focus on what changed. Possible causes include a new source system, software deployment, changed mapping, revised business rule, unusual batch, or manual processing error.
The current DMBOK2 Chapter 13 revision retains Statistical Process Control concepts while reorganizing them into supporting material and explicitly maps Shewhart/Deming improvement stages to Data Quality processes.
The signal does not prove what caused the defect, nor does it establish that performance has improved or deteriorated permanently. It indicates that the observed variation is sufficiently unusual to justify analysis.
Quality teams should preserve the evidence, correlate the timing with system and process changes, use lineage to identify affected sources, and investigate before adjusting the process unnecessarily.
Reference Topics: DAMA-DMBOK2 Chapter 13 — Statistical Process Control; Control Charts; Common Cause; Special Cause; Continuous Improvement.
===============
