How to Interview Employees After a Workplace Incident

Employee interviews are one of the most important parts of a workplace incident investigation. They are also one of the easiest parts to get wrong. Ask the wrong question and you may lead the witness toward your theory. Create the wrong atmosphere and employees may become defensive.

Wait too long and important details may fade. Treat the interview like an interrogation and you may learn less rather than more. A good incident interview is not primarily about getting someone to admit what went wrong. The goal is not to confirm your story.

The goal is to develop a better-supported one. OSHA guidance similarly recommends interviewing people promptly, using a fact-finding rather than fault-finding approach, reducing fear and anxiety, and avoiding leading or intimidating questions.

Interview People as Soon as Reasonably Possible

Memory changes. People discuss the event with one another. The worksite changes. Supervisors begin asking questions. Employees hear other versions of what happened. The longer you wait, the more difficult it can become to separate a person's original recollection from information they learned later. OSHA's employer investigation guidance recommends conducting interviews promptly after the site is safe and the immediate situation has settled. OSHA's enforcement procedures similarly emphasize identifying and interviewing people with firsthand knowledge as early as possible.

Prompt does not mean rushed.

People may need:

• medical attention,
• emotional recovery,
• emergency responsibilities,
• union or employee-representative involvement,
• language assistance, or
• time for the situation to become safe.

The goal is to obtain information while memories are relatively fresh without ignoring more immediate needs.

Interview People Separately

Whenever practical, interview witnesses individually. Separate interviews help preserve each person's independent recollection. If several employees reconstruct the event together first, their accounts may begin influencing one another. That does not mean disagreement between interviews is a problem. Differences can be useful. Two people standing in different locations may genuinely have seen different things.

An investigation should not force those observations into artificial agreement.

Instead ask:

Why might these accounts differ?

Position?

Visibility?

Timing?

Noise?

Attention?

Different information?

Those differences may help explain the event.

Explain the Purpose

Before asking detailed questions, explain why you are conducting the interview.

A simple explanation might be:

“I want to understand what happened and the conditions surrounding the work so we can learn from the event and prevent something similar.”

That establishes a different tone from:

“I need to determine who caused this.”

OSHA-supported investigation materials emphasize reducing fear and anxiety and explaining that the interview is intended for fact-finding and prevention rather than fault-finding. That distinction matters.

If employees believe the purpose is simply to find someone to blame, they may:

• become guarded,
• minimize uncertainty,
• tell investigators what they think management wants to hear,
• protect coworkers,
• protect themselves, or
• avoid describing normal work practices that differ from written procedures.

Learning depends on information. The interview environment affects whether that information becomes available.

Start Broad

Do not begin by feeding the witness your theory.

Start with something broad:

“Walk me through what happened.”

Then allow the person to talk.

Other useful openings include:

• “Tell me what you remember.”
• “What were you doing before the event?”
• “What happened next?”
• “What did you see?”
• “Where were you when this happened?”
• “How is this job normally performed?”

OSHA-supported training materials recommend open-ended questions and caution investigators against interrupting or leading witnesses. A broad opening gives the witness a chance to identify information the investigator may not know to ask about.

Do Not Interrupt Too Quickly

Investigators sometimes hear something important and immediately jump in. That can break the person's thought process. Allow the witness to finish the initial account. Take notes. Mark things you want to return to. Then clarify. Silence can also be useful. If the witness pauses, you do not always need to fill the space.

People sometimes remember additional details when given a moment to think.

Move From General to Specific

Once the employee has provided an uninterrupted account, begin clarifying.

For example:

“You said the load moved. What did you observe?”

“When you say the machine sounded different, what was different about the sound?”

“You mentioned that the area was busy. What was happening around you?”

“You said this was unusual. What normally happens?”

This approach lets the witness establish the initial account before the investigator begins narrowing the questions.

Ask About Normal Work

Do not interview only about the few seconds surrounding the incident. Ask how the work normally happens.

Questions might include:

• How is this task usually performed?
• What usually makes the task go smoothly?
• Where does the job normally become difficult?
• What changes from one job to another?
• What do experienced employees watch for?
• What do you normally do when this condition occurs?
• How closely does the written procedure match the actual work?
• Are there normal adjustments people make?

These questions help distinguish the event from the surrounding work system. That difference can be critical. An unusual event cannot be fully understood without knowing what “normal” looks like.

Ask What Was Different

One of the most useful questions is:

“What was different this time?”

Possibilities might include:

• different equipment,
• a different load,
• staffing changes,
• weather,
• lighting,
• schedule pressure,
• maintenance activity,
• another contractor,
• a new employee,
• changed traffic,
• production conditions,
• unusual noise, or
• different material.

You do not know the answer beforehand. That is precisely why the question is useful.

Ask What the Person Knew at the Time

Hindsight can make an event look obvious. The employee did not know the outcome while making decisions.

Ask:

• What were you expecting to happen?
• What information did you have?
• What could you see?
• What could you hear?
• What were you paying attention to?
• What did you think the equipment was going to do?
• What were you trying to accomplish?
• What options seemed available?

These questions help reconstruct the situation from the employee's perspective rather than from the investigator's post-event knowledge. Understanding why an action made sense at the time does not mean concluding that it was correct. It means trying to understand the decision accurately.

Avoid Leading Questions

A leading question suggests the answer.

For example:

“You were rushing because production was behind, weren't you?”

This question introduces both rushing and production pressure. The employee may agree even if those conditions were not central.

A better question is:

“What was happening with production at the time?”

Then:

“Did that affect the work in any way?”

Compare:

“You couldn't see the pedestrian because the load blocked your view, correct?”

with:

“What could you see from the operator position?”

The second question collects information. The first asks the witness to confirm your explanation. OSHA-supported training materials specifically warn investigators against leading questions.

Be Careful With “Why”

Why questions are useful. They can also sound accusatory.

Compare:

“Why did you do that?”

with:

“Walk me through what led to that decision.”

Both seek explanation. The second often produces more useful information.

Instead of:

“Why didn't you follow the procedure?”

consider:

“How is this task normally performed?”

Then:

“How did that compare with the procedure?”

Then:

“What influenced the way the task was performed that day?”

The objective is not to avoid difficult questions. It is to ask them in a way that improves the quality of the answer.

Separate Observation From Interpretation

Suppose a witness says:

“The operator wasn't paying attention.”

That is an interpretation.

Ask:

“What did you observe that led you to that conclusion?”

The witness may say:

“He was looking to the left while the forklift continued forward.”

Now you have an observation.

Another example:

“The supervisor was pressuring us.”

Ask:

“What did the supervisor say or do?”

This helps distinguish:

• what the witness observed,
• what the witness inferred, and
• what the investigator later concludes.

All three may matter. They should not be treated as the same kind of evidence.

Do Not Force Precision

People do not always remember exact:

• times,
• distances,
• speeds,
• sequences,
• words, or
• positions.

Avoid turning estimates into facts.

If someone says:

“It was probably about ten feet away.”

record it as an estimate unless other evidence establishes the distance.

If the person does not know, let the answer remain:

“I don't know.”

An honest uncertainty is better evidence than a guess produced because the investigator demanded an answer.

Ask About Controls

Interviews should explore what was supposed to prevent the event.

Ask questions such as:

• What controls were normally used?
• Which controls were in place?
• Were they working?
• Was anything difficult to use?
• Had the control caused problems before?
• Was there a way to bypass it?
• How did employees know the control was functioning?
• What would normally stop this event from happening?
• What prevented the outcome from being worse?

This connects the interview to the broader investigation. The objective is not merely to reconstruct human actions. It is to understand how the work system behaved.

Ask About Successful Work

Investigations naturally focus on failure.

Also ask:

“Why doesn't this normally happen?”

or:

“What usually keeps this job safe?”

Workers may describe:

• informal communication,
• sequencing,
• positioning,
• double-checks,
• equipment cues,
• experience,
• adjustments, or
• teamwork

that help the work succeed. Those practices may not appear in written procedures. They can reveal both strengths and vulnerabilities in the system.

Ask What Would Make the Work Better

Employees performing the task often have useful ideas.

Ask:

• What would make this job easier to perform safely?
• What would you change?
• Is there a better way to control this exposure?
• What have people suggested before?
• Is there something management should understand about this work?
• What could prevent a similar event?

OSHA-supported investigation guidance recommends asking interviewees what they believe could have prevented the incident while focusing on conditions and events preceding it. Do not assume every suggestion is automatically feasible. Treat it as information to evaluate.

Ask What You Are Missing

Near the end, ask:

“What haven't I asked you that I should have?”

or:

“Is there anything else I need to understand?”

This is one of the easiest ways to discover that the investigation has focused on the wrong issue.

You can also ask:

“Who else should I talk to?”

OSHA enforcement guidance emphasizes identifying all people with firsthand knowledge and recognizes that follow-up interviews may be necessary.

Re-Interview When Necessary

An interview is not necessarily a one-time event. New evidence may create new questions. A photograph may contradict part of the original timeline. Equipment data may identify a condition nobody knew about initially. Another witness may describe an important interaction. Return to the employee.

Say:

“I learned something new and want to understand how it fits with what you described earlier.”

OSHA's employer guidance expressly recognizes that investigators may need to re-interview witnesses as additional information becomes available. Updating an account as evidence develops is part of good investigation practice.

Confirm What You Heard

Before ending, summarize important points.

For example:

“Let me make sure I understand what you're telling me.”

Then describe the sequence or key facts.

Ask:

“Did I get that right?”

Allow corrections. This can prevent a misunderstanding from entering the investigation record and later being treated as evidence.

Document the Interview Carefully

Documentation requirements will depend on the organization, event, applicable policies, and legal circumstances.

At minimum, preserve:

• who was interviewed,
• their relationship to the event,
• when the interview occurred,
• key factual information,
• important uncertainties,
• relevant observations, and
• follow-up items.

If statements are recorded, signed, transcribed, or otherwise formally documented, follow applicable organizational, legal, labor-relations, and privacy requirements. OSHA's own enforcement investigations use formal witness-documentation practices, but an employer's internal investigation is not automatically governed by every procedure OSHA applies to its compliance officers. That distinction matters.

Do not accidentally present an OSHA enforcement procedure as a universal employer requirement.

Consider Language and Communication Needs

An employee should be able to understand the questions and communicate effectively. If language is a barrier, use appropriate translation or interpretation support. OSHA-supported incident-investigation material recommends conducting interviews in the employee's language and using a translator when needed.

Also consider:

• hearing limitations,
• literacy,
• technical terminology,
• stress,
• cultural differences, and
• communication style.

The purpose is to improve understanding—not to test how well someone speaks the investigator's language.

What About Representation?

Depending on the workplace, applicable law, collective-bargaining arrangements, company policy, and the nature of the interview, employees may have representation rights or other procedural protections. Do not improvise those requirements. Follow applicable labor, legal, and organizational guidance. For serious events, investigations involving potential discipline, fatalities, regulatory enforcement, litigation, or other high-consequence circumstances, coordinate with the appropriate organizational resources.

A learning-oriented interview still exists within a real legal and employment environment.

Do Not Promise Confidentiality You Cannot Guarantee

Employees may ask:

“Is this confidential?”

Be precise. Do not promise absolute confidentiality unless you can actually provide it.

Explain:

• how the information will be used,
• who may receive it,
• how the investigation process normally handles statements, and
• any limitations you know.

Trust is strengthened by accuracy, not by promises that may later be broken.

The Interview Is One Piece of Evidence

An employee statement should not automatically determine the conclusion.

Compare it with:

• physical evidence,
• photographs,
• video,
• electronic data,
• procedures,
• records,
• equipment condition,
• measurements,
• timelines, and
• other interviews.

People can be sincere and still:

• remember incorrectly,
• estimate inaccurately,
• miss something,
• misunderstand something, or
• interpret the same event differently.

That is normal. The investigator's job is to build the most supported account from the available evidence.

A Practical Interview Sequence

A simple structure can help.

1. Establish purpose

Explain that you are trying to understand the event and surrounding conditions.

2. Allow an uninterrupted account

“Walk me through what happened.”

3. Reconstruct the sequence

Clarify what happened before, during, and after the event.

4. Explore normal work

How is the job normally performed?

5. Explore differences

What was different this time?

6. Explore context

What could the employee see, hear, know, and expect?

7. Explore controls

What normally prevents the event?

8. Explore successful work

Why does this task usually succeed?

9. Ask for improvements

What could make the work safer or more reliable?

10. Ask what you missed

Who else should you talk to? What else matters?

11. Confirm your understanding

Summarize important points and allow corrections.

12. Follow up

Return if new evidence creates additional questions.

Questions Worth Keeping

Here are several useful questions to keep available:

“Walk me through what happened.”

“What happened immediately before that?”

“What happened next?”

“What could you see from where you were?”

“What were you expecting to happen?”

“What is normally different?”

“How is this job usually performed?”

“What usually makes this work go well?”

“What makes it difficult?”

“What controls do you normally rely on?”

“What was different this time?”

“What would make this work better?”

“What haven't I asked that I should have?”

Those questions will often produce more useful information than:

“Whose fault was it?”

The Standard: Learn Before You Conclude

Good interviewers are curious. They listen before deciding. They distinguish observation from interpretation. They allow uncertainty. They test what they hear against other evidence. Most importantly, they remember that the employee experienced the event without knowing how it would end. The investigator has the advantage—and the danger—of hindsight.

The interview should help reduce that distortion. Do not interview employees merely to confirm the explanation you already have. Interview them because they may know something that changes it.

For a closer look, read Root Cause Analysis vs. Learning From Events.

Supporting resources

Related commercial publications are disclosed under Authorship & disclosures below.