How Tim Page-Bottorff Thinks About Root-Cause Questions After an Incident

Episode 10 of Headline Podcast, published on November 19, 2025, features Tim Page-Bottorff, ASSP Senior Vice President, in conversation with Andreza Araujo and Dr. Megan Tranter.

His argument changes incident investigation by moving the first question away from the person who acted and toward the conditions that made the action possible, attractive, or difficult to challenge.

Why root-cause questions decide the quality of an investigation

A root-cause question determines whether an investigation produces a person to blame or a decision the organization can improve. In the first 24 hours after an incident, evidence is incomplete, memories are moving, and leaders are often under pressure to explain what happened before the work has been reconstructed. A useful question therefore has to preserve uncertainty while directing attention toward controls, planning, supervision, and the operating conditions that shaped the event.

Tim Page-Bottorff states the distinction plainly in Episode 10: “I don't think we should do a root-cause who. I think we should do a root-cause what.” The sentence is short, but it sets a demanding standard. Investigators still need to establish who made a decision, who owned a control, and who had authority to change the plan. They should not confuse accountability with the assumption that a person is the cause.

OSHA explains that an incident investigation should identify root causes and underlying causes, not stop at the immediate action or condition. That distinction makes the investigation useful before the organization decides which controls, resources, or routines must change.

What does “root-cause what” mean in practice?

“Root-cause what” means asking which condition, decision, control failure, or planning assumption allowed the event to develop, rather than treating the worker’s final action as the endpoint of the analysis. The question applies to a near miss, a serious injury, a process upset, and a weak signal that has not yet produced harm. In each case, the investigator follows the work backward until the system can explain why the risk remained available.

The approach does not remove human judgment from the investigation. It improves the target of that judgment. An operator may have chosen an unsafe route, but the team still needs to ask whether the safe route was visible, practical, resourced, supervised, and compatible with the production plan.

A practical sequence can move through five questions. What was the person trying to achieve? What information was available at the decision point? Which control was expected to prevent the deviation? What made that control weak, unavailable, or easy to bypass? Which owner can change the condition without relying on personal vigilance?

Those questions create a bridge between human action and organizational design. They also reduce the risk that a report will recommend retraining when the real correction belongs in equipment, sequencing, staffing, maintenance, or authorization.

How should investigators protect evidence before interpreting it?

Investigators protect evidence by separating observation, interpretation, and assumption before they write a causal conclusion. This separation matters because the first story told after an incident often becomes the story that survives, even when later records show a different sequence. A disciplined investigation establishes a time line, preserves original records, interviews people separately, and marks what remains uncertain.

Use the first hour to secure the scene, identify the people who were present, preserve relevant permits and logs, and record conditions that may change with cleanup or restart. Use the next 24 hours to compare statements with work orders, shift plans, alarms, training records, and control verifications. Do not turn a missing record into proof that a conversation never happened.

NIOSH publishes research and guidance that place work conditions, exposure pathways, and prevention controls at the center of occupational risk analysis. That lens supports a useful discipline for interviews. Ask what the person saw, expected, understood, and could do at the time, then test each answer against independent evidence.

Tim’s emphasis on discipline is relevant here. He said, “If you don't discipline yourself, someone else will.” In an investigation, that discipline means resisting the convenient explanation, naming the evidence that supports a claim, and recording the evidence that still needs to be tested.

Which questions expose pressure without turning pressure into an excuse?

Good questions expose pressure by showing how the plan, the schedule, and the available controls interacted before the incident. Pressure is not a complete cause by itself, and calling it an excuse can be as unhelpful as ignoring it. The investigator should identify the specific demand that changed the decision, the person who knew about that demand, and the control that should have absorbed it.

Ask whether the work was scheduled with enough time for isolation, testing, inspection, or handover. Ask what happened when the original plan no longer fit the conditions. Ask whether escalation was usable at the moment it mattered, or whether stopping work would have required a worker to absorb delay, embarrassment, or a production conflict alone.

A strong report distinguishes between a deliberate rejection of a workable control and a control that was unavailable, unclear, impractical, or inconsistent with the way the task was actually organized. Both situations require accountability, but the corrective decision is different. One may require a boundary or formal response. The other may require redesign.

ISO 45001 specifies a management-system approach in which hazards, risks, operational controls, competence, and improvement are connected. An investigation that names only the final behavior breaks that connection and leaves the organization unable to show how the system contributed.

What changes when the investigator asks “who” first?

Asking “who” first narrows the investigation before the evidence has been assembled. It can lead the team to search for a rule violation, a training gap, or a distracted worker while overlooking the conditions that made the deviation predictable. The result is often a familiar corrective action, such as a reminder, a briefing, or a revised poster, even though the hazard remains in the work.

The table is not an argument for ignoring responsibility. It is an argument for sequencing the inquiry correctly. Establish the conditions first, then decide which responsibilities follow from the evidence.

This is especially important when several people made reasonable decisions inside a weak system. A supervisor may have accepted a rushed handover because the staffing plan had already failed. A worker may have continued because the isolation point was not clearly identified. A manager may have known about the recurring problem without seeing a decision request that required action.

How can leaders turn the investigation into a control decision?

Leaders turn an investigation into a control decision when every causal finding is connected to an owner, a change in work, a verification method, and a date. A report is not complete because it contains five recommendations. It is complete when the organization can show which recommendation changed the exposure and how that change was tested in the field.

Start with the evidence chain. The evidence chain for decision-ready investigations helps distinguish a fact from an interpretation and a proposed action from a verified control. Use the first-24-hours evidence review when the incident is still being reconstructed, and compare the final method with the choice between Five Whys, Fishbone, and Fault Tree Analysis rather than selecting a method by habit.

Each finding should answer four operational questions. What must change? Who has authority to change it? What evidence will show that the change exists? When will the team check whether the change survived normal work, a production variation, and a shift handover?

Keep the review visible for at least 30 days when the corrective action changes a routine or a control. A completed action in a tracker is not the same as a control that workers can use when the task is difficult.

What should a supervisor do before the next investigation interview?

A supervisor should enter the next interview with a neutral reconstruction plan, not a preferred explanation. Before asking for a conclusion, write down the task objective, the expected controls, the point at which conditions changed, and the evidence that can confirm or challenge each part of the sequence. This preparation takes less than one hour when the records are accessible, yet it can prevent a rushed interview from defining the investigation.

Use open questions first, such as “What were you trying to achieve?” and “What did you expect to happen next?” Follow with specific checks about isolation, communication, equipment condition, staffing, time pressure, and authority to stop or re-sequence the work. Avoid asking questions that contain the answer, because they invite agreement rather than evidence.

After the interview, compare the account with at least three independent sources. A permit, a control-room log, and a work order may reveal different parts of the same sequence. The disagreement is not a nuisance to eliminate. It is a signal that the team needs to understand how information moved through the operation.

For a deeper comparison of evidence sources, the witness interview, document review, and field walkdown guide offers a useful decision frame. The aim is not to collect more material. It is to collect the right material for the decision the organization must make.

Recommendation

Use Tim Page-Bottorff’s “root-cause what” distinction as a review gate for every serious incident and high-potential near miss. Before approving the report, ask whether the team has identified the work condition, control, decision, or planning assumption that allowed the event to develop, and whether the named owner can change that condition.

Run the gate in three stages. Within one hour, preserve the scene and list the facts that can still change. Within 24 hours, compare the first account with records and separate evidence from interpretation. Within seven days, approve only corrective actions that define an owner, a verification method, and a date for field confirmation.

Keep the question visible in the investigation template: “What allowed this to happen?” Then add the accountability question after the evidence is clear: “Who has the authority and duty to change it?” That order keeps the report fair without making it soft, and it keeps learning connected to a decision.

Conclusion

Tim Page-Bottorff’s Episode 10 argument offers a practical correction to weak incident investigations. A root-cause “who” stops at the person. A root-cause “what” follows the conditions, controls, information, and decisions that made the event possible.

Listen to the full conversation with Tim Page-Bottorff on Episode 10, then test the question in your next investigation before the first explanation hardens into the official story.