skills.rest
jcg-admin
jcg-adminCOMMUNITY
Explore all skills by this author →
GitHub

pdca-check

Compare PDCA pilot results against SMART objectives and validate statistical significance.

Updated Mar 14, 2026
Search Tags:#root cause analysis#continuous improvement#statistical significance#pdca#pilot evaluation#smart objectives#run chart
One-click install
npx skills add https://github.com/jcg-admin/IACT-ui --skill pdca-check-jcg-admin
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pdca-check
Source: https://github.com/jcg-admin/IACT-ui/tree/main/.claude/skills/pdca-check
Command: npx skills add https://github.com/jcg-admin/IACT-ui --skill pdca-check-jcg-admin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? After running a PDCA pilot, teams often declare success or failure based on gut feeling rather than data. This Skill structures the Check phase of the PDCA cycle, comparing actual pilot results against the baseline and SMART objectives defined in the Plan phase, and producing a data-backed verdict. ## Core Features & Use Cases - Objective vs. Result Comparison: Builds a metrics table contrasting baseline, SMART target, and actual pilot outcome, including control metrics to detect regressions. - Statistical Significance Validation: Applies Run Chart rules (8 consecutive points), sample-size minimums, T-test, Mann-Whitney U, or historical range checks to distinguish real improvement from noise. - Structured Verdict and Recommendation: Produces a direct conclusion (hypothesis confirmed, refuted, or mixed) plus a recommendation for the Act phase: standardize, adjust, retry, or revert. - Use Case: After a two-week pilot reducing API p95 latency, use this Skill to verify the sample size is sufficient, confirm the improvement exceeds natural process variation, and generate the pdca-check.md artifact with a clear verdict. ## Quick Start Ask the assistant to run the PDCA Check step on your completed pilot data and compare the results against the Plan objectives to produce a verdict.

Frequently Asked Questions about pdca-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if a PDCA pilot actually improved the process?▼

Compare pilot results against both the baseline and the SMART objective from the Plan phase, then verify statistical significance. A Run Chart with 8 or more consecutive points on one side of the pre-change median indicates a real signal rather than random noise.

What sample size is needed to conclude a PDCA pilot?▼

It depends on process frequency: continuous processes need at least 30 post-change observations, intermittent processes need about 4-6 weeks, and low-frequency processes need at least 3 complete cycles. Insufficient samples only support a preliminary verdict.

When should I use a Run Chart versus a T-test?▼

Use a Run Chart whenever you have time-ordered data, regardless of sample size. Add a T-test only when both groups have 30 or more observations and roughly normal distributions; otherwise prefer Mann-Whitney U or the historical range check.

What happens if pilot data is insufficient for a conclusion?▼

The Check step cannot conclude with confidence when the sample is too small or the pilot too short. Options are to extend the Do phase, or document the Check as preliminary and schedule a later review.

Can I adjust the objective after seeing pilot results?▼

No. The objective is fixed in the Plan phase; changing it after seeing results invalidates the Check. If results fall short, the honest verdict is partial improvement, leading to an adjusted plan in the next cycle.

Related Skills

Explore More Product & Management Skills→
jcg-adminjcg-admin

pdca-do

Executes the Do phase of PDCA by running controlled pilots and collecting measurement data.

Community
Intermediate
NestorMonroyNestorMonroy

pdca-act

Decides whether to standardize or adjust a PDCA improvement and documents lessons learned.

Community
Intermediate
tunnckoCoretunnckoCore

six-sigma

Apply DMAIC and DMADV frameworks to improve AI agent processes.

Community
Advanced
jcg-adminjcg-admin

dmaic-control

Create Control Plans, SPC charts, and reaction plans to sustain DMAIC improvements.

Community
Advanced