As a Health and Safety Inspector for many years, I have interviewed witnesses to serious events, including fatalities to workers and members of the public. As Human Factors Manager in a major oil and gas company, I was involved in investigations of health, safety, quality and productivity events. Key to all of these investigations was the way that an interview was conducted. Besides gathering documentation and inspecting the site of an event, speaking to witnesses is a key source of information in understanding what happened and why.
The investigation of incidents requires a wide range of technical and non-technical skills. Every investigation should include at least one team member with experience of previous investigations.
In this article I will outline how to conduct an interview following an adverse event, such as an incident or near-miss. I provide a general approach to asking questions and getting the most out of the time that you may have for the interview. Other articles provide suggestions on how to prepare for an interview and the actual questions to ask in an interview. The aim of this article is to simply provide a framework for asking these questions.
Why interview a witness?
The aim of an interview is to help a witness recall everything they know in their own words, without being influenced by either the question, or by what they think investigators want to hear. Note that I use the term ‘witness’ quite widely – it refers not just to someone who observed an event, or who was nearby at the time, but all of those individuals who can provide information relating to the investigation. This may include the Leaders of the organisation, those responsible for key processes and systems, as well as the designers of the workplace or equipment. Therefore, a witness does not need to have been present at the time of an event.
How not to interview a witness
The traditional approach to interviews by police, physicians, safety inspectors etc, is to ask specific questions directed toward each fact that the interviewer needs to learn. The interview may begin with demographic information, followed by an open-ended question such as “What happened?”. However, typically, within a few seconds, the interviewer interrupts and asks a rapid-fire of closed questions. When the relevant factors have been exhausted, the interviewer closes with “Is there anything else that you wish to add?”. This type of approach, although common, fails to elicit all of the information – and may be detrimental to the psychological health of a witness.
“Let me ask the questions
and you give me the answers”
In these interviews, the interviewer dominates the discussion, and the witness plays a small role. The focus is on a set of facts or evidence that the interviewer requires. The interviewer does most of the talking – and the sequence of the interview is determined by the interviewer (often driven by a checklist of pre-determined questions). The questions themselves are usually forced-choice and may be suggestive or leading. At worst, the questions are used to confirm a hypothesis (pre-determined idea) about the event.
Information processing 101
We know that people have limited mental resources to process information.
So, witnesses may not be able to listen to an interviewer’s questions AND search through their memories at the same time. Therefore, to prevent a witness from being overloaded, it is good practice to refrain from asking more questions whilst the interviewee is thinking how to answer a questions, or trying to remember something.
I’ve written about working memory elsewhere.
Overall approach
If there are several witnesses they should be interviewed one at a time. If there is more than one interviewer, assign roles beforehand. Ideally, one team member should ask the questions and another take notes. Towards the end of the interview, the person asking questions can invite other team members to ask questions. This approach helps the Lead Interviewer to maintain their line of thought, and prevents the witness from being bombarded with questions.
If I am the one asking questions, I sometimes invite other team members to ask questions at certain points when I want to pause and gather my thoughts, or make a note of something important. The role of questioning then returns back to me when I am ready. I always agree my preferred approach with other team members before the interview.
The following approach to interviewing is recommended:
- Ask the interviewee to tell what happened from their point of view, from start to finish. Allow them to speak freely without any interruptions or prompts. It may be useful to make encouraging noises, or nod your head etc. to show that you are listening. Keep an open mind, do not judge, and be a good listener!
- Focus on critical junctures in the sequence of events – the interviewee can assist in determining these.
- Progressively probe and build a picture of the sequence of events and how the world looked to the interviewee at each juncture. It may be helpful to refer to maps, photographs, sketches and documents at this stage.
The “cognitive interview”
This approach to interviewing leads to interviewees giving much more accurate and relevant information compared with a traditional interview method. It originates from work by Geiselman & Fisher to improve police interview techniques, and is based on applied psychology. This work has shown that interview techniques can be improved considerably through training. Many police and safety investigation agencies around the world now use this style of investigative interviewing. The approach has been validated experimentally and in real-world applications.
The “cognitive interview”, as their work was described, is based on two principles of memory:
- A memory is composed of several elements. The more elements a memory-jogging aid has in common with the memory, the more effective the aid is likely to be.
- A memory has several different ways of being retrieved, so information that is not accessible with one method may be accessible with a different one.
Key techniques in a cognitive interview
The aim is to get the witness to start talking with minimal intervention from the interviewer (known as ‘free recall’). This is a great way to start an interview. Do not interrupt a witness talking through their free recall – not even for clarification! Just let them tell the story in their own words.
The following approaches have been found to greatly increase the amount and accuracy of information gathered in an interview. They will help the investigation team to build an accurate picture of what happened.
1. Change the perspective: Ask the interviewee to recall the situation from another person’s point of view. For example, from the view of a bystander, another witness, or perhaps an injured person. What might they have seen from their perspective? For example, if the interviewee was a bystander to a road accident, what might they have seen as the driver of a vehicle involved in the event?
2. Recreate the context: Instruct the interviewee to recall the environmental features and the way they felt during the event. In this approach, the interviewer is attempting to reinstate the environmental and personal context of the event for the witnesses, perhaps by asking them about their general activities and feelings on the day. This could include sights, sounds, feelings and emotions, or the weather. By mentally taking the witness back to the scene of the event, it is possible to improve their recollection of key factors. This has a similar positive effect on the interviewee as placing them physically back where the event occurred. Ask them to relax and recall where and when the event occurred. They should then be invited to look around it in their mind and note who was present, what they could see, hear, touch and smell. The aim here is not just to get the witness to describe events, but to re-live them.
3. Change the sequence: Ask the interviewee to recall the event in a different narrative order rather than chronological order. In this approach, the interviewee may start their explanation at the end and work backwards to the beginning. Or they may be asked to start in the middle of their description.
4. Report every detail: Instruct the interviewee to report every detail from the event, regardless of whether they consider it relevant. Even trivial or apparently unimportant details are requested; as they may trigger key information about the event. Interviews may require people to provide more detail than they would do in normal conversation. If a high level of detail is required from the interview, this should be made explicit. Interviewees should be encouraged to report everything, even if it is out of sequence or contradicts something said earlier in the interview.
5. Tailor the questions: Each interviewee has a unique mental record of an event – they will have focussed on different aspects of a situation. Tailor your questions towards what someone has focussed on. Rather than asking a standard set of questions in the same order for each witness; ask questions that are relevant to (1) what the witness may have focussed on and (2) what the witness is currently thinking about. Even during an interview, the details that a witness can recall will vary in their accessibility from memory. Being sensitive to the mental focus of the witness, and timing your questions accordingly, will lead to more effective recall. Geiselman & Fisher called this “witness-compatible questioning” and it requires the interviewer to structure an interview around the recollections of the witness, rather than proceeding in a pre-determined sequence.
6. Multiple retrieval: Psychology studies have shown that the more often we search our memories about a particular event, the more new details we will recall. Therefore, it is good practice to ask a witness to describe a critical event several times during an interview. Depending upon the severity of the event, and the role of the witness, it may be necessary to interview them several times. As witnesses may continue to think about an event after the interview, a follow-up phone call may also lead to new details being recalled.
The cognitive interview may not feel like an interview at all.
The interviewer asks few questions, and the witness appears to be just telling a story.
Avoiding bias
It is possible for an interviewer to introduce a bias in the responses of a witness by asking leading questions. Leading questions can usually be answered by Yes or No. For example, a leading question might be “Was the car red?”.
It is preferable to commence with open-ended questions, such as “What can you tell me about the car?”. Open questions tend to lead to a longer, richer, narrative response from a witness – and also have the benefit of making them feel more in control of the interview.
Open questions usually start with the words ‘what’, ‘who’, ‘when’, ‘why’, ‘where’ or ‘how’. You may also find it useful to create questions based on the acronym TED (“Tell me . . .”, “Explain to me . . .”, “Describe to me . . .”).
Once the witness has provided information, they can be asked follow-up closed questions, such as “What colour was the car?”. Closed questions should be used sparingly, for example when a complete response has not been provided to an open question.
Benefits for the witness
Although we might understand the benefits of the cognitive interview in terms of accuracy and amount of information gathered, this style of interview also has many positive effects for the witness. They are more in control, feel that they have ‘told their story’ and feel less like they have been interrogated. Because open questions allow witnesses to almost always provide some information, they may feel more successful than in conventional interviews, where they may be unable to answer some of the specific questions. The ability to talk more and control the interview enables witnesses to feel that they performed well in the interview.
When asked a series of closed-questions, witnesses may feel like suspects. However, in the cognitive interview, they are permitted to give open-ended narratives. The style, pace and social dynamics of the cognitive interview are more likely to support better psychological health of witnesses.
Want to know more?
12 questions to ask in an investigation. The purpose of these questions is simply to encourage getting behind the label of ‘human error’ and to better understand the ‘why’.
Human error, human performance and investigations. Human error is a central concept in ergonomics and human factors. But what is ‘human error’? Is it helping us to improve safety?
Fisher, R. P. & Geiselman, R. E. (1992) Memory-enhancing techniques in investigative interviewing: The cognitive interview. Springfield, IL: C.C. Thomas.
You must be logged in to post a comment.