Eyewitness testimony (EWT) is a legal term for the account of an event, given by a witness or observer. This could include specifics about a suspect, victim, the timeline of events or perhaps details about the scene. The testimony of a witness is powerful evidence for a jury, who seem to put a lot of confidence in EWT. If you think about it, by the time a case gets to trial, a witness may have repeated their story over and over until they are 100% confident in their memory of the event. Retelling this rehearsed testimony in court would be hard for a jury to doubt.
And yet, as I mentioned in my Police Lineups 101 post, over 70% of overturned wrongful convictions are the result of faulty eyewitness identification.
What are some of the reasons EWT might be incorrect? And how might you include some of these factors in your story?
Seeing a weapon during an event absorbs our focus because it is something that is unexpected and frightens us. Witnesses who observe a weapon during an event are therefore able to accurately recall details about the weapon they saw, but less so identifying characteristics about the suspect.
Johnson and Scott ’76 had two groups of participants sit in a waiting room. Group one heard an argument in a nearby room and a man carrying a pen ran out. Group two heard an argument in a nearby room and a man carrying a blood-covered knife ran out. When asked to identify the man holding the item, group one were able to identify the man with 49% accuracy whereas for the group who observed the knife, only 33% accuracy was recorded. This supports the idea of weapon focus – witnesses are distracted by the weapon and so are less able to remember what the suspect looked like.
Our brains have a lot to processes at once, and in busy environments such as witnessing an attack, it must be selective in what it focuses on. It makes sense then that the brain would choose to focus on the aspect of its environment most likely to cause harm: the weapon.
Pozzulo et al ’07 discovered an effect akin to weapon focus in shy witnesses. People who are shy tend to remember details about the suspect but less so about the event and surroundings. This works like weapon focus in that the shy person feels threatened by a new person entering their environment and so that new person draws their focus.
Is a weapon observed by your witness? How able are they to describe the suspect?
Up to a point, stress increases our cognitive performance, but then it drops off. Like a bell curve. This means that, under moderate stress we may be able to remember lots of detail about an event we witness, but if that event becomes very stressful, for example if a gun is produced, we may then remember much less.
Valentine and Mesout ’08 demonstrated the effects of anxiety on witness recall in a study of two groups at London Dungeon. 45 minutes after experiencing a frightening encounter with a live actor in the dungeon, participants were asked to identify the actor. Those that had experienced a high anxiety state during the encounter were able to identify the actor with 17% accuracy. Those that experienced only a low anxious state during the encounter were able to identify the actor with 75% accuracy.
We can reliably conclude then that experiencing high levels of anxiety during an encounter makes you less accurate at identifying a suspect later on.
How well can your witness recall the event they saw? Did they feel stressed during the event or calm?
When we are trying to remember an event we have witnessed, our brains fill in any gaps with something it thinks is logical. This is a bit like playing a game of Chinese whispers – if we are unsure exactly what we have heard, we take a ‘best guess’. The problem with reconstructive memory is that what each person deems a ‘logical’ gap filler, will change based on that persons cultural background and norms, biases and upbringing.
To use an example, a witness may retell a story of seeing a woman with long blonde hair run from the scene of a crime, but can’t remember what the face looked like. It may well be that the witness did not know 100% whether the suspect was a man or woman, but assumed it must be female based on norms in Western society; men are less likely to have long blonde hair than women.
Is this a man or a woman? Can you be 100% sure?
Engstler-Schooler ’90 concluded that it is in fact a very difficult task to verbalise a description of a face. It is no wonder then that our brains try to help us out by adjusting what we see so that it is easier to remember.
Does your witnesses’ story include additional details to help them make sense of what they saw?
Our brains are very susceptible to false memories. This was displayed in the 1990s when psychotherapy was under fire for ‘helping’ patients remember false memories such as satanic rituals and even pregnancy – you can find out more about this in a TED talk by Elizabeth Lofus:
These are questions that by their very nature, suggest a desired answer to the witness. Loftus and Palmer ’74 did a study of groups of participants observing a video of a car crash, who were then asked to estimate the speed the cars were travelling. One group were asked how fast the cars were going when they hit each other, the other were asked how fast the cars were going when they smashed into each other. The second group reported faster speeds, around 40.8mph compared to just 34mph for the group told that the cars ‘hit’ each other. ‘Smashed’ is a more emotive word and is suggestive of more damage than ‘hit’.
It is therefore imperative that during police questioning, words are chosen carefully so as not to lead the witness into altering their recall of an event.
This could be a conversation between witnesses of an event, seeing a news report or even a conversation between an eyewitness and interviewer after an event that contaminates the witnesses’ recall. This is why it is so important to separate witnesses as soon as possible upon identifying a crime scene, so they cannot corroborate stories.
Does your witness change their story after speaking to someone else? Does your police officer deliberately mislead the witness to satisfy their own agenda?
Unsurprisingly, alcohol has negative effects on the memory. This is a problem because many crimes are witnessed under the influence of alcohol or drugs, calling into question the validity of eyewitness testimony from drunk bystanders.
Oorsuow and Merckelbach ’11 put 76 participants into 3 groups: zero, moderate and high intoxication levels. They were shown a video of a crime and asked about it under interview several days later. Compared to sober participants, highly intoxicated participants remembered up to 33% fewer details about the crime.
Interestingly, subsequent studies have shown little difference in recall between sober and highly intoxicated participants so the jury is out on whether this factor significantly affects the validity of eyewitness testimony. The key cause here may be the repetition of interviews after the event to ensure all details are extracted from the witness.
Is your witness intoxicated? Are they able to remember all details clearly? Does it take more than one interview for them to remember? Are they treated differently by police during the interview?
Retention of memories declines over time. This means that the more time that passes in between witnessing an event and retelling it, the worse that recall will be, perhaps forgetting whole pieces of the event. This is known as the ‘forgetting curve’. It is therefore imperative that witnesses are identified and interviewed as soon as possible following an event.
Time can also affect witness testimony in terms of exposure to the suspect. As you can imagine, studies support the idea that longer exposure to a suspect during an event means better recall during a witness statement.
How much time passes between witnessing the event and giving the testimony? Are any significant details forgotten?
Very young and old witness’s testimonies are less reliable than young to middle aged witnesses. Bartlett and Leslie ’86 suggest that this is because people are better at remembering ‘same-age faces’ as opposed to those younger or older than themselves. Given that most suspects are not children or elderly, it makes sense then that these groups of people may struggle to identify a middle-aged suspect.
Searcy et al ’99 studied two age groups who watched a video of a crime and were then asked to identify the suspect from a photo lineup. The first group were aged 18 to 30 years, and the second aged 60 to 80 years. They found that the older group made more false choices during identification.
Given that Western society is ageing, finding reliable witnesses may therefore become harder for criminal investigations.
Children’s brains are not fully developed, for example displaying limited memory capacity and recognition skill compared to adults. Children are also highly suggestible, especially to misleading information. Research by Ceci and Bruck ’93 found that many factors influence a child’s ability to give eyewitness testimony including peer pressure, wanting to impress authority figures and overactive imaginations.
Many studies have found that juries view children as less credible witnesses than adults. Goodman et al ’87 found for example that the credibility assigned by a jury to the evidence given by a six-year-old was less than that assigned when it was given by a ten-year-old, which in turn was less than that assigned to a thirty-year-old.
How old is your witness and the suspect? Is the witness able to accurately identify the suspect? If there are age differences, what might the witness get wrong? Do they make any assumptions about the suspect?
Studies suggest that women make more reliable eyewitnesses than men, particularly with regard to accuracy of person descriptions and describing victims. Males however, are thought to be more accurate in describing the event itself.
‘Own-gender bias’ has also been proven in studies, meaning witnesses can more accurately identify faces that match their own gender; we can assume then that we may be less accurate at identifying a suspect who is the opposite gender to us.
How well does your witness fare in identifying a suspect from the opposite gender? Perhaps a couple witness a crime and are interviewed separately – do they remember different aspects of the event?
The problem of cross-racial bias exists; people struggle to differentiate accurately between races other than their own.
Goldstein and Chance ’85 suggested that people struggle to identify ‘other race’ faces they are not familiar with or have not been exposed to often. This skill is however something that can be improved upon over time as people are more exposed to other races. This could suggest that eyewitnesses who live in culturally rich locations such as capital cities would make more reliable witnesses than those from perhaps rural, culturally non-diverse locations.
Do differences in race affect your witness’s ability to recall the event and identify the suspect?
It goes without saying that ability to recall details about a crime would be significantly reduced under environmental factors such as poor/low light, nightime versus daytime, adverse weather conditions like fog, heavy rain, sleet etc.
One study by Forgas et al ’09 actually found that weather-induced bad mood makes for better accuracy of memory recall. This could mean that an eyewitness to a crime who is in a bad mood due to adverse weather may actually remember much more of the event!
Vallano and Compo ’11 found that the nature of the interviewing environment can make a big difference to eyewitnesses. In a study where interviewers took the time to build a rapport with witnesses, the accuracy of the accounts increased and they found that the witnesses were less likely to be affected by misleading information.
What time of day does your crime take place? What is the weather like? Do these factors affect the crime or ability to witness it clearly? What efforts do the police make to ensure the witness is comfortable?
Over to you
Consider each of the 10 factors and the questions posed. Do any of these feature in your story?