Five Whys vs Fishbone vs Fault Tree: Which Method Fits the Evidence?

A serious incident review can become less reliable when the team chooses a familiar method before it understands the evidence. Five Whys, Fishbone, and Fault Tree Analysis are not interchangeable templates, because each one exposes a different relationship between the event, its conditions, and the barriers that should have prevented harm.

The right choice depends on the question the investigation must answer. Five Whys is useful when one causal chain needs disciplined testing. Fishbone is useful when several contributing conditions may be interacting. Fault Tree Analysis is strongest when the review begins with a defined top event and must test how combinations of failures could produce it.

Use Five Whys for a focused causal chain, Fishbone for a broad contribution map, and Fault Tree Analysis for a serious top event whose combinations of failures must be examined. The method should follow the evidence, not replace it.

Evaluation criteria for choosing an investigation method

The first criterion is the shape of the evidence. If the event has a relatively clear sequence and the team needs to test why a condition persisted, a chain-based method can keep the discussion precise. If the facts come from different people, shifts, interfaces, and work conditions, the investigation needs a structure that allows those contributions to remain visible before anyone decides which ones matter most.

The second criterion is consequence. A minor process deviation may need a proportionate review, while a potential serious-injury or fatality event deserves a method that can examine barrier combinations and latent conditions. James Reason's work on organizational accidents is useful here, because it shows why the person closest to the event is only one part of the causal picture.

The third criterion is the decision that must follow. An investigation should lead to a change in design, authorization, staffing, supervision, maintenance, training, or escalation. Andreza Araujo's Safety Culture: From Theory to Practice makes the same practical point from a culture perspective, since repeated decisions reveal what the organization actually protects.

Before selecting a method, ask what the review must make visible, what evidence is still missing, and which operational leader has authority to change the condition. Those answers usually narrow the choice faster than a debate about which tool is most sophisticated.

Five Whys is best for a focused causal chain

Five Whys works when the team can state a specific problem and move backward through a sequence of conditions without treating the first plausible explanation as the answer. The method is not a ritual of asking the word why five times. It is a discipline for testing whether each answer is supported by evidence and whether another condition sits behind it.

Consider a dropped load that was released before the exclusion zone was clear. A weak review might stop at operator inattention. A stronger chain asks why the release was possible, whether the signal was visible, who had authority to stop the lift, whether the plan reflected the actual layout, and how supervision verified the critical step. Each answer should lead to a condition that can be checked in records, interviews, equipment, or the worksite.

The method becomes unreliable when the team forces every event into one line of causation. Human decisions are shaped by equipment design, time pressure, handover quality, planning, and local norms, which means a single chain can hide parallel contributors. Use Five Whys after the evidence has been scoped, not as a shortcut that closes the inquiry early.

Choose it when one failure sequence is dominant, the event boundary is clear, and the corrective decision can be tied to a small number of verified conditions. Pair it with a field check so the final answer does not remain a narrative that sounds logical but leaves the exposure unchanged.

Fishbone is best for multiple contributing conditions

Fishbone analysis gives an investigation team a visible way to organize possible contributors across people, equipment, materials, methods, environment, management, and other categories that fit the operation. Its value is breadth. It prevents the first interview or the most visible error from becoming the entire explanation.

That breadth matters after a near miss involving a temporary repair. The repair itself may be relevant, but the review also needs to examine how the defect was detected, whether the spare part was available, how the temporary measure was approved, whether the risk assessment was updated, and what the next shift understood about the condition. A Fishbone map keeps those threads together while the team separates evidence from assumption.

The risk is an unfiltered brainstorm. A diagram full of possible causes is not an investigation result, because possibility is not proof. For each branch, record the evidence that supports it, the evidence that weakens it, the owner who can verify it, and the decision that would change if the branch were confirmed.

Fishbone is the better starting point when the event crosses departments, shifts, contractors, or work phases. It is also useful when the organization has a history of explaining incidents through individual behavior while overlooking planning, resource, or design conditions. The final report should narrow the map into verified contributors rather than publish the entire brainstorming surface.

Fault Tree Analysis is best for a defined top event

Fault Tree Analysis starts with a top event, such as a person entering an energized zone or a pressure boundary failing, and works downward through combinations of conditions that could make that event possible. Logical gates help the team test whether the top event requires one failure, several failures together, or an earlier condition that defeats more than one barrier.

This structure is especially valuable when the potential consequence is severe and the operation depends on layered protection. The review can examine isolation, authorization, verification, physical separation, alarms, human response, and recovery arrangements without assuming that one missing step explains the whole exposure. It also creates a clearer bridge between incident evidence and critical-control assurance.

Fault Tree Analysis demands a precise top event and enough technical evidence to define the relationships between failures. If the team cannot agree on what happened or what the barriers were meant to do, the tree can create false precision. The diagram may look rigorous while the underlying assumptions remain untested.

Choose it when the review concerns high-consequence potential, barrier interaction, or a system in which several independent failures may align. A control owner and an operations leader should review the final tree together, because EHS can challenge the logic but operations owns the changes that make the barrier dependable.

Decision matrix: match the method to the evidence

The comparison becomes practical when the team uses the same criteria for all three methods. The table below is a starting point for an EHS manager deciding how much structure the review needs.

Use the least complicated method that can answer the decision honestly, but do not use simplicity as an excuse to ignore serious exposure. An investigation of a low-consequence process error may need a focused chain. A high-energy event with multiple barriers may justify a tree even when the diagram takes longer to build.

In more than 250 cultural transformation projects supported by Andreza Araujo, the practical distinction is visible in what happens after the report. If the method produces a polished explanation but no change in work design, authority, or control verification, the organization has completed documentation rather than reduced risk.

Common errors that weaken all three methods

The first error is starting with blame. A person may have made the final action, yet the investigation still needs to examine the conditions that shaped that action, including planning, equipment, workload, competence, supervision, and local signals about what mattered most.

The second error is treating a completed corrective action as proof of effectiveness. A revised procedure or retraining record shows that something was done. It does not show that the barrier is available, understood, used as intended, and effective when conditions change. That distinction is also central to the four investigation myths that make root cause analysis look complete.

The third error is losing the original evidence. Photographs, statements, equipment condition, shift records, and change history should be secured early, because a causal method cannot recover facts that were allowed to disappear. The incident evidence handoff process is useful when multiple teams need access to the same material without changing its traceability.

The fourth error is separating the incident review from serious-risk governance. When a potential fatal exposure appears, the investigation should connect with the review of critical controls rather than remain a local story. The analysis of SIF investigation distortions shows why severity potential must remain visible even when the outcome was less severe.

Recommendation: choose the method that changes the next decision

Start with Five Whys when one causal sequence is clear and evidence can test each link. Start with Fishbone when the event crosses interfaces and the team needs to keep several contribution paths visible. Start with Fault Tree Analysis when a defined top event and interacting barriers require system logic.

Then define the next decision before the report is approved. The decision might change a guard, isolation step, staffing level, contractor interface, approval right, maintenance priority, or field-verification cadence. Name the operational owner, state the evidence that will confirm the change, and set the date on which the organization will test whether the exposure actually moved.

Andreza Araujo's work in safety leadership keeps this choice grounded in action. Her books, including Make The Difference: Be a Leader in Health & Safety, are useful resources for leaders who need investigations to strengthen daily decisions rather than only improve the language of reports. Safety is about coming home, and a method earns its place when it helps the operation protect that outcome.