Root Cause Analysis Checklist — Free PDF Download

Root Cause Analysis Checklist

This checklist covers the 8D D4 stage. Its purpose is simple: make sure an investigation does not stop at the first plausible answer. A good root cause analysis explains the physical mechanism, the system gap that allowed it, and why the problem reached the customer.

Use it before you sign off D4, and keep the evidence attached to each line.

Pre-Analysis Preparation

• The problem is defined using 5W2H (what, where, when, who, which, how, how many)
• The cross-functional team is assembled (D1)
• Containment actions are in place and verified (D3)
• All relevant data is collected: measurements, photos, records, operator statements

Why this matters: a root cause analysis run on a vague problem definition produces a vague fix. The 5W2H statement is the foundation of every Why that follows.

A practical test before you start: hand the 5W2H statement to someone outside the team. If they cannot describe the defect, the location and the quantity back to you without asking a question, the definition is not ready. Containment must also be verified, not just planned — confirm with a count of contained stock and a check that the containment station is actually catching parts, otherwise you will be investigating a moving target.

Root Cause Tools Checklist

• 5-Why analysis completed to at least 5 levels
• Fishbone (Ishikawa) diagram covering all 6M categories
• Is/Is-Not analysis used to narrow the scope
• Fault Tree Analysis (FTA) applied for complex multi-factor causes
• Escape point identified — why did the defect reach the customer?

Tool notes:

• Fishbone first, then 5-Why. The fishbone widens the search across Man, Machine, Method, Material, Measurement and Environment; the 5-Why then drills into the most likely branch.
• Is/Is-Not tests what the problem is and is not, which prevents fixing a cause that only explains part of the failure.
• FTA is worth the effort when several factors combine, or when safety is involved.

Worked Example: Bore Diameter Out of Tolerance

A machined housing shows a bore diameter of 10.08 mm against a specification of 10.00 ± 0.05 mm on 6 of 200 parts from one shift. The 5W2H statement fixes the defect (oversize bore), the location (line 2, OP-40), the time (night shift, after tool change) and the quantity (6 of 200).

Is/Is-Not narrows the field: the defect is on parts made after a tool change, is not on parts from the previous tool, is on the second half of the batch, is not on the morning shift. The fishbone branch that survives is Machine/Method — tool wear compensation. The 5-Why chain reads: bore is oversize → the tool cut beyond nominal → the offset was set from a warm-up part → the offset procedure does not specify a stabilized temperature → the work instruction was never updated when the machine was replaced (MRC). The escape point: the first-piece inspection passed because it was taken from the warm-up part, and the in-process gauge was checked only at start of shift.

That example shows the shape every D4 should take: a measured mechanism, a documented system gap, and a detection gap with a named cause.

Step-by-Step Verification Sequence

1. Re-read the 5W2H statement and confirm it still matches the data collected.
2. Walk each Why answer and attach the evidence file or measurement to it.
3. Reproduce the failure in a controlled test — change one variable at a time.
4. Reverse the test: restore the original condition and confirm the failure disappears.
5. Separate TRC from MRC and write each as a single sentence.
6. Trace the escape point through the detection controls that should have caught it.
7. Have a team member who did not run the analysis challenge the chain.
8. Record the D4 sign-off with the evidence index attached.

Common Mistakes

• Stopping at the first plausible cause. Verify with data before you commit.
• Blame instead of mechanism. "Operator error" is a starting point, not a root cause.
• One linear chain for a complex problem. Use the fishbone to find competing causes first.
• Ignoring the escape point. Even a good root cause fix leaves the detection gap open.
• No reproduction. If you cannot make the failure happen on demand, you have not proved the cause.

How to detect these early: if every Why answer is a sentence with no number, photo or record behind it, you are writing opinion. If the chain reaches a person's name or a training gap within three levels, the mechanism was skipped. If the Is/Is-Not table is empty, the scope was never tested. If the reproduction step is missing from the evidence pack, treat D4 as open regardless of how convincing the narrative reads.

How D4 Links to the Neighbouring Steps

D4 does not stand alone. It consumes the D2 problem description and the D3 containment record, and it feeds D5 and D6 directly. A weak D4 shows up later as a corrective action that does not hold: D5 selects permanent actions from the verified causes, so an unproven cause produces a countermeasure aimed at the wrong variable. D6 then validates effectiveness against the same measurements used in the reproduction test, which is why those measurements must be recorded at D4. The escape point feeds D6 as well — the detection control change is verified separately from the cause fix. Finally, D7 reads the MRC to decide which documents, PFMEAs and work instructions need updating.

What a Complete D4 Looks Like

A complete D4 answers four questions: what physically failed (direct cause), why it happened (TRC), why the system allowed it (MRC), and why it was not caught (escape point). If any of the four is missing, the analysis is not finished.

→ Generate your root cause analysis automatically with the free AI 8D Report Generator at report.8d.wiki