Home > Attention, Memory > Are Eyewitness Accounts Trustworthy?!

Are Eyewitness Accounts Trustworthy?!


Ever wonder how accurate eyewitness accounts actually are?  Picture yourself sitting in a room conversing with police officers about a crime that you witnessed earlier that day.  Will you be able to remember every detail of the event and people involved?  Was all of your attention focused on the crime and no other factors?  How can we be sure that every memory of the event is even true?  Eyewitness accounts in general have been a topic of concern for these very reasons.  Many defendants are proven guilty based on eyewitness accounts; some of which consist of distorted/untrue statements.  To read more about this topic, click here.   Due to these issues, there is a growing interest in researching eyewitness testimonies in psychology, especially in the cognitive field.

Memory in general is far from being perfect.  Errors are very common in memory.  There are different types of errors that can occur, one category being when introducing information to memory. Occasionally memories aren’t even formed because information fails to be encoded and stored.  Errors can also be made when trying to retrieve information from memory.  A widely known example of this is the concept of forgetting, where information in memory fades over time.  Can you recall every detail and event that occurred on last Tuesday?  Probably not, which demonstrates how quickly information can be forgotten, making it more important to have testimonies given very soon after the crime and to not wait many days.  As the elapsed time between the event and recall increases, the probability of forgetting increases with it.  Memories can also change by having details be added or removed, all happening unconsciously.  We, as humans, don’t have total control on what we remember and what information we forget. Because of these flaws in our memory, eyewitness accounts are proving to be less accurate.

Errors commonly occur in times of high stress because people don’t have to ability to pay full attention to everything occurring.  We can choose what to pay attention to, but the more items we try to focus on, the less likely all of the information will be stored in memory.  In cases like this, such as crime scenes, it is not specific details that are always stored, rather a summary of the situation.

Eyewitness testimonies are never going to be completely removed from our judicial system because they play crucial roles and are occasionally the only source of information about the event that is accessible to the police and jury.  Thus why psychologists have grown interested in finding the most reliable type of eyewitness testimony, either being written or spoken.  When taking an eyewitness account, police can either ask for a written report or can conduct a personal dialogue about the series of events.


Speaking and writing put different demands on the cognitive system.  The type of account that requires fewer demands allows the eyewitnesses to fully distribute their resources to the memory task.  There are different lines of research that show superiority for speaking.  An advantage is that it is faster, allowing for quicker recall and lessening the amount of time information is stored in memory; therefore limiting the amount of forgetting.  Another advantage is the use of body language and tone fluctuations when in an interview.  These factors can give away hints about their confidence in the information they are recalling.  In previous research it has been found that spoken accounts are more accurate, but display more falsehoods (Kellogg, 2007).  However, writing allows for self pacing and censoring over formerly produced information.  Therefore leading to fewer errors than speaking (Horowitz & Newman, 1964).  Writing also has the ability to be more efficient when there is numerous eyewitnesses because they can all give their accounts simultaneously, consequently reducing the rate of forgetting.

In 2014 a group of five psychologists assembled and created a study testing the accuracy of written accounts versus spoken accounts of eyewitnesses.  In the study the participants watched a short video of a staged crime between two women.  They were told to watch carefully because they had to report on it later.

The two different conditions that were manipulated in the experiment were the type of account the participant had to give and whether or not an interviewer was present.  It is important to see if the presence of an interviewer influences the accuracy of the reports.  This is because interviewers have the ability to guide eyewitnesses and prompt them to retrieve more details.  The cues that interviewers may give are crucial in the reports witnesses give.

In the first experiment, participants were told to report everything they could remember as entirely and correctly as possible without guessing.  However, in the second experiment, participants were asked to recall everything they remembered.  Basically the first group was given stricter instructions, which were supposed to help the participants’ recollections.  The experimenters hypothesized that the more specific guidance would aid the participants to recall more relevant information.  In both experiments there were general and specific questions asked about the crime and the people involved.  The experimenters also hypothesized that the more detailed instructions would be most beneficial on the more demanding task.

In general for both types of account, it was found that the stricter instructions led to more specified accounts; meaning that directions given to participants play a crucial role in the retrieval of the targeted information.  Also, for both experiments, the written accounts were more detailed, but the accuracy of every detail is questionable.  Contrasting to one of the hypotheses mentioned, the participants who gave oral reports did not benefit from the more detailed instructions, although the written reporters did.  One possible explanation for this result is that the participants who had to speak might have had a hard time recalling the instructions, where as the instructions were in front of the writing participants.  It was also found that the participants who wrote their reports did better in the absence of an interviewer.  This could be due to the absence of the extra stress of someone watching over them.  In conclusion, there was no significant evidence supporting whether written or spoken accounts are more accurate.  So far it has been concluded that both writing and spoken eyewitness accounts appear to be suitable.

A way to facilitate the most accurate eyewitness recollections needs to be discovered to help make the judicial system trust worthier!

To read more about Eyewitness Testimonies, try the following links:

Faulty Eyewitness Testimony

Innocent Criminal


Horowitz, M. W., & Newman, J. B. (1964). Spoken and written expression: an experimental analysis. Journal of Abnormal and Social Psychology, 68, 640-647.

Kellogg, R. T. (2007). Are written and spoken recall of text equivalent? American Journal of psychology, 120,415-428.

Sauerland, M., Krix, A.C., van Kan, N., Glunz, N. & Sak, A. (2014). Speaking is silver, writing is golden? The role of cognitive and social factors in written versus spoken witness accounts. Memory & Cognition. Retrieved from http://0-link.springer.com.library.colby.edu/article/10.3758/s13421-014-0401-6/fulltext.html

  1. Emily Doyle
    October 7th, 2014 at 23:24 | #1

    Fun article choice! I enjoyed reading about eyewitness accounts because they are so important to our legal system as a whole. The difference between vocal and written reports was interesting because it was inconclusive, both types considered suitable, with pros and cons to both sides. I found it quite important that the outcome of the study was actually more dependent on whether the interviewer was present or not or how strict the instructions were, because they could cue the response in some way, or stress the interviewee out, and the instructions acted as a cue of what to retrieve. While reading this article it struck me though, no matter whether written or vocal accounts are more accurate, the most important part of an accurate eyewitness account is how well the eyewitness remembers the account in general. Recently in my cognitive psychology class we learned about the seven sins of memory, or really seven memory malfunctions. These malfunctions explain losing memory or remembering incorrect memories. For an eyewitness to report on a crime, they first have to properly remember the crime. If any of these malfunctions are at play, then the crime may be remembered differently no matter what the form of response is. One sin of memory is the sin of suggestibility, where a person can become confused in their recollection of an event based off of comments made by others. I think this sin comes into play in this article because the interviewer can lead the witness in their questioning or by body language, this can actually modify the memory of the eyewitness as they integrate this misleading information into their account. While this was not investigated in this article, I think that it is an important aspect of this sort of memory retrieval and should be looked into more in the future.

  2. October 6th, 2014 at 13:57 | #2

    This study brings up some interesting points regarding eyewitnesses. The Capacity Framework Model proposed by Kahneman supports your idea that a witness’s attention is limited when he is in a stressful situation because stress requires mental capacity, which means the crime witness has less available mental capacity to attend to the crime scene. This inability for a witness to direct his full attention to the scene makes his account of the scene questionable in terms of how accurate it is. In this study, I like how researchers focus on ways to make the accounts more or less accurate by testing how well they could help the witness retrace their memory of the crime. For example, the researchers tested witnesses to see if they would provide more detailed and accurate accounts of the scene when given strict/demanding instructions. It makes sense that the witnesses who wrote down their accounts of the scene were able to provide more detailed analyses of what they saw when given stricter instructions because these instructions strengthened the memory trace; there must have been words or times within the questions that helped the witnesses formulate more detailed accounts–the witnesses could rely more on top-down processing. On the other hand, it also makes sense that witnesses who gave oral reports did not give more detailed accounts for stricter instructions because an oral report limits one’s capacity. When the witness listens to the question, he must remember the question being asked while attending to the information from the scene stored in his long-term memory. Again, the Capacity Framework Model supports this notion that when a person uses mental energy towards one task, such as remembering a question being asked, he has less mental energy to attend to another task, such as retrieving information regarding the crime scene.
    Although this study is helpful because it shows that we can assist witnesses retrace their memories of a crime scene by asking strict questions, it does not help determine how accurate these accounts are. It would be interesting to use this technique of asking strict questions to help measure the accuracy of these detailed accounts. One way to test accuracy could be to test witnesses’ accounts of crime scenes that consist of people of same vs different races or ages than the witness. According to the Harrison and Hole article regarding own-age and own-race bias, a person have the tendency to better-recognize people of his own age and his own race than people of different ages or races. This ability to better-recognize people more like yourself relies on attention; you pay more attention to people who look more like you and you pay less attention to people who look less like you. Could you test this for a crime scene? So, if a 20 year-old Caucasian witness sees a crime between a 20 year-old Caucasion person and an Asian child, than the witness will pay more attention to the 20 year-old; therefore, the witness will better-recognize the 20 year-old. Because the witness pays more attention to one person in the scene, his account of the 20 year-old should be more accurate than his account of the child.
    Because a witness’s account varies depending on the witness himself and who is in the crime scene, it is unfair to ask the general question: “Are eyewitness accounts trustworthy?” Rather, some accounts will be more trustworthy than others. Perhaps researchers should take into account the race and age of the witnesses and relate it to the people in the crime scene to test how accuracy varies across the accounts of the different people in the scene. If witnesses of the same age and race of a person in the crime scene can more accurately identify that person than a witness of another age/race, then the witness with the more accurate account should be trusted more for his memory of that particular person in the scene.

  3. Caroline Minott
    May 10th, 2014 at 01:13 | #3

    With numerous empirical studies already establishing eyewitness testimony as an unreliable source in the legal system, no matter how meta-cognitively “confident” the witness is in their story, I really enjoyed how Sauerland and her colleagues dug deeper into eyewitness testimony reliability by comparing the different testimony formats. I just spent this spring semester collaboratively carrying out a PS 215 study on the misleading nature of written eyewitness testimony, so I was especially interested in reading about spoken eyewitness testimony. While this study examined the accuracy when generating eyewitness testimony, our study looked at how already-generated written eyewitness testimony statements can impact a co-witness’s recollection of a crime. Specifically, we investigated how reading an accurate vs. inaccurate eyewitness statement of a petty crime influences blame attribution and memory accuracy for the events/details of the crime. Just in the same way that this study manipulated the format of the eyewitness testimony, I wonder whether we would’ve saw the same accuracy and blame attributions if someone was asked to recall crime details after reading vs. listening to an eyewitness testimony. Spoken language adds a new layer of complexity, with speech being formed through an endless combination of different phonetic features. Although we categorically perceive language in a way that allows us to comprehend meaning even with speaker variations in articulation and volume, we’ve learned in class that stress patterns in speech can alter the meaning. In the same way, it’s obvious that emotionally stressed phrases or words also alter interpreted meaning in eyewitness testimony.

You must be logged in to post a comment.