Fishbone analysis
Map possible causes, then investigate them with evidence.
- Collect possible causes
- Do shift patterns or experience levels correspond to the defects?
- Collect possible causes
- Are defects concentrated on a machine or after a setup change?
- Collect possible causes
- Does quality differ between material batches?
- Collect possible causes
- Do procedural differences coincide with defects?
- Collect possible causes
- Do environmental changes align with the affected batches?
- Collect possible causes
- Does the finding reflect a product change or a measurement error?
These six categories suit a manufacturing example and can be adapted. A connection is not proof of causation: clarify, consolidate and test candidates with evidence.