If you witnessed a car accident and were later asked to explain what you saw, what words would you use to illustrate the scene? Think of all the different terms or phrases we could use to describe the impact. Did the cars crash? Did they collide? Or maybe they bumped into each other, or one hit the other? As we identify what word to use to describe the accident, it’s often related to things like the perceived speed or the amount of damage we observed. An investigator may even ask several witnesses to describe the event, and each provides a different term which relates to their perspective of what they observed.
What causes this? Besides our perspectives or experience with other similar events (maybe we’ve seen accidents at much higher impact in our past than the other witnesses in comparison), there are a lot of variables that may actually cause a person’s memory to be altered when they are asked to recall an event. Research shows that even the way in which an investigator asks the question could alter a person’s memory. In fact, if the investigator asked, “how fast were the cars going when they hit each other?” versus “how fast were the cars going when they crashed?” it has been shown to alter the way in which a person remembers the event.
In this example, witnesses who were primed with the “hit” question may remember a slower speed and less damage than those who received the “crash” version. This is a primary example of how investigators may contaminate a person’s memory simply by the way we ask a question. Additionally, with both questions starting with the phrase “how fast” the investigator is already suggesting that speed was a factor in the accident. It is possible that a witness may consciously align with the investigator’s perspective in their effort to help the investigation. It is also likely that a witness may unknowingly provide inaccurate details due to the suggestive and leading questioning.
We know that most of you are not conducting traffic accident investigations, but this example of witness contamination translates across employee relations issues, fraud, theft, or any other fact-gathering conversation you may find yourself in. As we illustrate some of the key reminders below, we challenge you to think of how each premise relates to your specific investigations. Organized retail crime investigators often require conversations with store personnel to explain what they observed; those details can be essential in developing a case file for prosecution. Employee relations issues like workplace violence, threats, discrimination, or other inappropriate behavior are often reliant on an investigator’s ability to obtain accounts from witnesses. If we alter a witness’s memory and then rely on those details to make further case decisions, the foundation of our investigation is built on a fallacy.
Using Open-Ended Questions
This may be the simplest of advice, but it is often the most ignored. Asking a witness about our car accident in an open-ended format allows them to choose a word with free will, versus any implicit suggestion or contamination. “Could you tell me what you observed?” does not provide the witness any subjective terminology in which they may align their story with. This also allows the witness to start the story where their memory is most readily available, versus being forced to draw out a specific detail or in any order. These intricate pieces of information can be explored with further questioning, based on the witness’s initial response.
In a workplace violence situation where two employees were observed in an altercation, these same rules may apply. If the interviewer asks, “how aggressive was Dave in that argument?” there is a suggestion that Dave was aggressive, and that the conversation made it to the level of “argument.” These subjective terms may suggest something different from a witness’s perspective, resulting in a contaminated recall of events. In these investigations, especially relying on the credibility of witness statements, this could mean the difference between a termination of an employee or not.
Another common occurrence in the loss prevention field is the fact-gathering interviews that take place with store personnel after an organized retail crime incident. If an investigator learns of a violent shoplifting event or organized crime interaction at a location, they will most likely ask store personnel about what they observed. An open-ended question like, “could you tell me about the experience yesterday in the store?” will yield more information with less contamination. However, in these cases it is possible that the investigator already has some information about the presumed suspects. This known information can be dangerous when attempting to question witnesses as it may result in leaked details. A common error could be initially asking the store personnel something like “was it a group of big guys wearing jackets?” or “did this group all have airpods in when they walked in the store?” Although this may be crucial information, it would be more reliable if a witness can provide those details unprompted.
Avoid Confirmatory Feedback
When a witness provides information, it is helpful to encourage memory retrieval and thank them for their cooperation throughout. However, investigators must be careful not to confirm a witness’s recall as accurate or consistent with other evidence. When an investigator makes a confirmatory statement, such as “that makes sense, we thought it was an older male with glasses” they are increasing the confidence of the witness’s memory. This increased confidence is not a result of the witness’s actual recollection but rather the confirmation by the interviewer that they are “on the right track.”
If we think about this concept from an eyewitness perspective, we can see how easily it is for investigators to contaminate a person’s memory through confirmation. Using our organized retail crime example, an investigator may have several pictures of known shoplifters involved in an ORC ring that they show to a store associate. The witness may be unsure of the exact match, but because they want to help the investigator, they point at a picture of an individual that they believe looks like the person who committed the theft in their store. The investigator’s reaction to this identification can be dangerously impactful, as a simple statement like “I figured it was him” or “that’s what we thought” will provide confirmation to the witness they got it right. If this investigation moves forward and law enforcement interviews the store associate to obtain an affidavit of what they observed, we are now dealing with a contaminated memory. The store personnel will likely identify with the same picture they chose earlier, but this may be a result of the confirmation they got it right versus a reliance on what they actually saw.
Post-Event Discussion
Reflecting back to our car accident example, it’s possible that multiple witnesses observed this incident from the sidewalk or their own vehicles. Naturally, as witnesses wait for first responders to arrive, there is going to be discussion between them. One witness might share how they couldn’t believe how fast one of the vehicles was traveling, and another may recall how one of the vehicles cut them off earlier. Another witness might interject themselves stating that they saw the light was red for at least five seconds before one of the vehicles entered the intersection. These post-event interactions can easily contaminate a witness’s memory, altering their version of events to align with their peers. It may also serve as confirmation of their own (potentially inaccurate) memory. This same phenomenon happens to all of us when meeting with a group of friends or family members who are recalling some event from years ago. As the story gets retold, and multiple people add their contributions, the resulting memory may be far from the truth—yet consistent among the involved parties.
In an ideal setting, investigators should attempt to isolate witnesses and interview them in close proximity to when the incident occurs to prevent post-incident discussions. These types of situations could be prevalent in the loss prevention industry anytime multiple witnesses are interviewed about an event. This can prove difficult for investigators when assessing credibility, as these statements may be very consistent with each other—but not as a result of accurate recall, rather a product of post- event discussion. Investigators should be aware of this error and include this within their interview process, asking witnesses if they have talked to anyone or have seen any other coverage of the incident prior to the conversation.
There are a variety of other variables that may contaminate or alter a person’s memory. Investigators should be aware of things like dispositional risk factors, exposure to trauma, or the emotional state of the individual at the time of the event (and the time of recall). Exposure to media coverage or other publicly available information may contaminate memory. Interview methods, such as the Cognitive Interview, have been shown to increase memory recall while mitigating the risk for contamination. This is generally done through effective open-ended questioning, the use of silence, and allowing the interviewee to guide the conversation.
For more information on this topic in the field of memory research, check out the work of Elizabeth Loftus, PhD and John Palmer, PhD starting with their foundational paper that provided the context and data in the “car crash study.” “Reconstruction of Automobile Destruction: An Example of the Interaction between Language and Memory.” Journal of Verbal Learning and Verbal Behavior 13.5 (1974): 585-89.