Structured and semi-structured interviews help safety professionals understand how people experience a system, task or working environment. They can reveal hazards, weak controls, conflicting responsibilities and differences between documented procedures and actual practice.
The method provides individual stakeholder input to risk assessment. It is useful when people cannot attend a workshop or may be reluctant to speak openly in a group. The findings need careful interpretation: an interview records an account, not automatic proof that a control works or a risk estimate is correct.
Choose the interview format for the purpose
A structured interview uses a prepared question set, normally with consistent wording and order. This helps the team compare responses about the same topics. Clarification can still be necessary, but the interviewer should record any significant departure from the agreed format.
A semi-structured interview uses a common topic guide and allows more flexible follow-up. The interviewer can explore an unexpected concern, ask for an example or investigate why an apparently simple task is difficult.
Use either format to identify risks, examine controls or understand stakeholder concerns. A structured approach supports consistency. A semi-structured approach provides more scope to discover issues that the original questions did not anticipate.
Define the objective and select participants
Write down what the interviews must establish. For example, the objective might be to understand how maintenance staff confirm electrical isolation, or how operators respond when an alarm is unreliable. Define the equipment, tasks and operating conditions in scope.
Select people with relevant experience across roles, shifts, locations and organisational boundaries. Include contractors and affected users where appropriate. Do not assume that the most senior or most available people provide a complete picture.
Interviews can reach more stakeholders than one small workshop, but the selection still matters. A set of convenient volunteers does not establish that all relevant experiences are represented. Record the roles covered, gaps and reasons for the selection.
Prepare neutral questions
Use simple language and ask about one issue at a time. Start with an open question that allows the person to explain their experience. Follow with specific questions to clarify the sequence, conditions or supporting evidence.
Instead of asking whether a procedure is clear and easy to follow, ask the person to describe a recent task. Then explore where they needed clarification. Avoid questions that imply the expected answer or assume that a failure occurred.
The GOV.UK Service Manual recommends preparing topic guides, testing questions and using open, neutral prompts. It also encourages discussion of real experiences. These interview practices transfer well to safety work. [1] Useful prompts include:
Please describe the last time you performed this task.
What information did you use to decide it was safe to start?
What made the task difficult or different from the procedure?
How do you know that this control is available and effective?
What happens if the normal method cannot be used?
Prepare follow-up prompts, such as asking what happened next or requesting an example. Test the guide with a colleague before the main interviews. Revise unclear wording without steering participants towards a preferred explanation.
Create conditions for a useful conversation
Explain the purpose, how responses will be used and who will receive the findings. Agree arrangements for notes or recording. Describe confidentiality accurately; do not promise anonymity if a person’s role or example will identify them.
Give the participant time to think and answer. Listen without interrupting to defend a procedure or explain the official position. Summarize important points and check that you have understood them. In a semi-structured interview, follow relevant new issues before returning to the guide.
Record the person’s account separately from the interviewer’s interpretation. Note whether a statement describes direct experience, an expectation or something heard from another person. Use established arrangements to escalate an immediate safety concern.
Test the account against other evidence
Interviewees can recall events differently or provide the answer they believe is expected. Interviewers can introduce bias through wording, tone, selective follow-up or interpretation. A common guide and review of notes can reduce inconsistency, but cannot remove these effects.
Compare responses across participants while preserving important differences. Different answers may reflect different shifts, equipment configurations or task conditions. Do not treat disagreement as a reason to discard a concern.
Check significant claims against observations, maintenance records, drawings, procedures and test evidence. If a person says an isolation valve is always accessible, inspect the location and consider conditions during the relevant emergency. Agreement between interviewees is useful information, but it is not a substitute for verification.
Case study interviews before the Texas City disaster
The US Chemical Safety Board’s final report describes a safety culture assessment at BP Texas City in late 2004. The Telos Group surveyed 1,080 employees and interviewed 112 people. A report issued in January 2005 identified serious concerns about production pressure, training, equipment conditions and resources. The refinery explosion on 23 March 2005 killed 15 workers and injured 180 people. [2]
This case shows that interviews can reveal important warning signs before an accident. The practical lesson is to connect findings to assessment, decisions and completed action. Collecting stakeholder views is only part of that process.
The CSB account establishes that interviews formed part of the assessment. It does not establish that every interview followed a particular structured or semi-structured protocol.
Produce findings that support decisions
Summarize themes, specific scenarios, differences of view and evidence gaps. Retain enough context to explain who is affected and under what conditions. Use participant identifiers where needed to support traceability while limiting unnecessary personal detail.
Assign actions for unresolved hazards or uncertain controls. Identify the owner, required evidence and decision authority. Link significant findings to the Hazard Log and relevant risk assessment.
Interviews take time to arrange, conduct and analyze. They also lack some of the idea-building that occurs in a group. Use them alongside observation, document review or a later workshop where those methods help. Their main contribution is access to individual experience that the wider safety assessment might otherwise miss.
Hashtags
#SafetyInterviews, #RiskAssessment, #HazardIdentification, #HumanFactors, #SystemSafety, #StakeholderEngagement, #SafetyAssurance
Sources
[1] GOV.UK Service Manual. Using in-depth interviews. Interview planning, question design and conduct. Read source
[2] US Chemical Safety Board. Refinery Explosion and Fire BP Texas City. Report 2005-04-I-TX. March 2007. Section 9.4.16 Telos Survey, printed pages 173 to 176; incident summary. Read source
