Module 4 · Lesson 2 of 22
Process-door and data-door analysis
Learning objectives
- Combine process-door and data-door analysis.
- Sequence qualitative and statistical evidence.
- Avoid blind fishing in software.
Two doors, one truth
Process-door tools (gemba, maps, Five Whys, C&E, C&E matrix, FMEA) generate causes. Data-door tools (stratification, plots, tests, models) test them.
Convergence between the two doors is the strongest routine confidence Analyze can provide.
Garment-factory example
Gemba observation flags folder alignment; stratified defect data confirm the effect only appears on one fabric construction and speed band; a return to gemba confirms the mechanism.
Method
- Generate candidate causes at the process door.
- Turn each into a data-testable prediction.
- Test with the appropriate data-door method.
- Return to gemba to confirm the physical mechanism.
Common mistakes
- Data fishing without a process theory.
- Accepting expert opinion without any data check.
Knowledge check
Pick one answer per question. Explanations appear after you submit.
1. The strongest routine Analyze confidence comes from:
2. Data fishing without a process theory tends to:
Author: Sanjeewa Dehiwalage · Last reviewed: 2026-07-21