This is the category-agnostic shelf — the pillar guide that covers the core RCA methods, the side-by-side comparisons for picking between them, the adjacent frameworks (PDCA, CAP, Fault Tree, 8D), and the fundamentals that every good analysis rests on (problem statements, corrective actions, curated reading).
Most RCA writing on the internet picks one method (usually 5 Whys) and pretends the others don’t exist. In practice, no single method survives every problem — you pick based on whether you need depth or breadth, reactive or proactive work, qualitative or data-driven prioritisation, and whether a regulator will read the output.
This category holds the cross-cutting material: a pillar guide that walks through the 6-step RCA process and the five major methods; four head-to-head comparisons; adjacent frameworks like PDCA and Fault Tree Analysis; and the two fundamentals that quietly determine whether any RCA works — a sharp problem statement and a corrective-action plan that actually ships.
If you’re choosing which method to use, start with the tools comparison. If you want the full playbook, read the complete guide.
Grouped by purpose: the pillar guide and comparisons at the top, adjacent frameworks in the middle, fundamentals and reading at the end.
What RCA is, the 6-step process, five methods compared, and how to turn findings into lasting fixes that actually ship.
Comparison5 Whys, Fishbone, Pareto, FMEA, Fault Tree — inputs, outputs, time cost, team size, and regulatory fit.
ComparisonDepth vs breadth, qualitative vs structured. Decision matrix and the common mistake of picking the wrong one.
ComparisonWhen a hunch beats a chart, when a chart beats a hunch, and how to run them together.
ComparisonQualitative brainstorm vs quantitative top-down probability tree. Safety-critical vs process-improvement contexts.
MethodBoolean gate logic, cut sets, and probability calculation for aerospace, nuclear, and medical-device applications.
MethodDeming's four-phase improvement loop, how it wraps any RCA method, and the trap of skipping Check.
FundamentalsThe 5W+1H frame, what to strip out, and why vague problem statements doom every RCA that follows.
ClosureThe CAP template, SMART actions, verification vs validation, and how most corrective actions die before they ship.
ReadingTwenty-three curated titles by Ishikawa, Deming, Ohno, Tague, and others. Every book has been used in actual practice.
RCA is the practice of finding the systemic cause of a problem — not the first plausible answer, not the individual who happened to be on shift, and not the symptom — so that a corrective action can prevent recurrence rather than just suppress this instance. The complete guide walks through the full 6-step process.
Depends on three questions: Is the failure known or hypothetical? Do you have data or only opinions? Do you need regulatory defensibility? See the tools comparison for the decision matrix — it covers all five methods side by side on inputs, outputs, time cost, team size, and regulatory fit.
Yes, and usually you should. Common pairings: Fishbone + 5 Whys (map breadth, drill depth), Pareto + 5 Whys (rank, then investigate), FMEA + Corrective Action Plan (score, then act). Running any single method in isolation is the most common source of shallow RCA output.
PDCA is the outer loop — Plan-Do-Check-Act is the improvement cycle that contains your RCA. RCA sits inside the Plan phase. Check is where you verify that the corrective action from your RCA actually reduced the problem (this is the step most teams skip). See the PDCA article for the detailed workflow.
SMART (specific, measurable, assigned, relevant, time-bound) + a verification step + a systemic target (process change, not “remind people to be careful”). Most corrective actions die because they have none of these. The Corrective Action Plan article walks through the template.
The recommended books list covers 23 curated titles: Ishikawa, Deming, Ohno, Tague, Roser, and others. Every book has been used in actual consulting practice — no one-star Amazon listicles.
Three guided RCA tools, no signup, no tracking, nothing leaves your browser. Start with 5 Whys if the cause is known, Fishbone if it isn’t, Pareto if you have the data.