Analyze submission pack

Learning objectives

• Assemble an Analyze submission pack that confirms vital causes.
• Combine process-door and data-door evidence.
• Reject Pareto height or p<0.05 as sole root-cause proof.

Required evidence

• Cause inventory from process-door and data-door
• Fishbone and Five Whys
• C&E matrix and process FMEA
• Stratified graphs
• Appropriate tests or models with assumption checks
• Effect sizes and confidence intervals
• Mechanism narrative and confirmed vital-X table

Copy-friendly Analyze pack checklist

• Each candidate cause has an evidence question
• Alternatives and confounders tested
• Findings triangulated across sources
• Causes classified: confirmed / unsupported / inconclusive
• Vital X list has temporal logic and credible mechanism

Garment-factory example

A seam-damage analysis considers fabric lot, needle type/life, SPI, machine model, method, shift and inspector. After stratification and paired tests, needle life and SPI are confirmed as vital Xs; machine model is unsupported.

Method

1. Link every candidate cause to a specific question.
2. Test alternatives before naming a root cause.
3. Publish the inconclusive list, not only the confirmed one.

Common mistakes

• Treating the tallest Pareto bar as the root cause.
• Data mining without process logic.
• Designing solutions before confirmation.