Academy / Chapter 02 / Method selection

Choose the method.Not the favorite.

Start with the question your evidence can answer. Techniques organize thinking, but they do not turn assumptions into facts or create data you do not have.

Academy overview →Open glossary ↗

QUESTION FIRST_

Use the problem shape to select a tool.

Choose the row closest to your immediate decision. Each technique can be combined with the others; it is not a linear ladder or a maturity ranking.

Question you actually have	Useful starting method	Failure to avoid
Which categories create the most recurring impact? You have multiple observations using comparable measurements.	Pareto Analysis Sort count, downtime or cost consistently.	The tallest bar identifies a priority, not a causal explanation.
What might be causing this effect? Several competing explanations remain plausible.	Fishbone / Ishikawa Map people, process, technology, environment and more.	Brainstorming creates hypotheses, not evidence.
Why did this specific chain occur? One supported causal sequence deserves deeper examination.	5 Whys Ask why each link exists and verify the transitions.	Five is not mandatory; branching problems need other tools.
What is different about a failure that occurs only sometimes? Specific IS/IS NOT contrasts and changes are available.	Kepner–Tregoe Problem Analysis Compare the deviation, distinctions and changes.	A distinction suggests where to test; it does not certify a cause.
Which response or investment best meets our needs? Several alternatives have competing risks and benefits.	Kepner–Tregoe Decision Analysis Separate MUSTs and weighted WANTs.	A numerical ranking is advisory; it does not replace judgment.

WHEN TO LEAVE THIS TOOLSET_

For a safety-critical event, medical/aviation process, destructive engineering test, high-uncertainty statistical inference or regulated incident, use the applicable specialist methodology, risk controls and qualified practitioners. The four tools shown here are introductions—not a universal root-cause solution.

INTELLIGENT HANDOFFS_

A method rarely works alone.

Pareto → Fishbone

Counts reveal that 61% of delays relate to picking errors. Use Fishbone to map plausible reasons the picking process produces errors.

PRIORITIZE → MAP

Fishbone → 5 Whys

Evidence elevates an unverified calibration branch. Test the branch, then trace the supported control failure with a causal chain.

MAP → EXPLAIN

Problem Analysis → Decision Analysis

Once the deviation is understood well enough, compare mitigation alternatives against non-negotiable constraints, benefit and risk.

DIAGNOSE → CHOOSE

QUALITY GATES_

Five questions before you present a conclusion.

Observable: Is the problem statement a deviation rather than an accusation?
Comparative: What is unaffected, or what happened before the change?
Traceable: Can another reviewer find the original logs, records, measurements and dates?
Falsifiable: What would you expect to see if your favored cause were wrong?
Actionable: Can the selected control be tested and the result monitored against a baseline?

Next / Chapter 03: Professional Case Files

See the method handoffs in four fictional operational investigations.

Study the cases →