How Memory Fails: Exploring the Unreliability of Human Memory

The concept of memory is, effectively, denoted to “mental time travel”. The formation of memory is comprised of three variating sub-processes: encoding, consolidation and retrieval of the constructed material. Individually, these processes are vulnerable to specific errors, particularly within misleading post event information and stress inducing events, subsequently creating a false memory. Memory is utilised to recall information each day, by every person on the planet; what you ate, past conversations, eyewitness testimony, and even the construction of temporal belonging. Considering the importance of human memory as a source of information within society, it is essential to determine whether society can rely on its the accuracy. This essay will argue that human memory, in relation to the details of past events, is not reliable.

There is an overabundance of observed phenomena for the existence of falsified or altered memory, including: biased recall, where the individual perception of the event distorts recollection; misinformation effect, specifying the negative consequences of exposure to misleading information prior to recollection; and memory conformity, which denotes recollection of an event can be influenced by another person’s recollection. For the purposes of this essay, however, the thesis will be supported through the interpretation of the empirical evidence collected by the joint journal of Morgan, Southwick, Steffian, Hazlett and Loftus and Marsh and Greenberg. The former demonstrates that memories for stressful events are highly vulnerable to modification by exposure to misinformation. The latter reports are consistent with the assumption that memory bias allows people to reduce threats through blaming the victim of a crime for not taking advantage of opportunity to avoid victimisation. Both scenarios relate to the real-life implications found within the jury process and reliability of eyewitness testimony.

Analysis of the First Study

As denoted previously, the research conducted by Marsh and Greenberg proposes that the motivation to reduce one’s stress can produce distortions in memory. In order to investigate this, the study examined the impact of victim to eye witness physical similarity and the victim’s culpability on eyewitness’ recall of time and distance. The study utilised a total of 187 participants: 90 women and 97 men, all Caucasian and ranging in age from 18-24. All participants were undergraduates enrolled in introductory psychology courses and were participating to receive course credit. In this particular study, there were two independent variables:

1. physical similarity between participant-observers and the victim (similar, dissimilar);
2. victim culpability (culpable, non-culpable).

The dependant variables in this study were the memory for distance and time – where participants were asked to provide estimates of the minimum and maximum distance between the criminal and the victim and an approximate of the duration of the incident.

The stimuli consisted of video tapes of four versions of a criminal victimisation scenario, where the two independent variables were manipulated. The participants viewed a one-minute video of a simulated street robbery. The physical similarity was manipulated by presenting a victim of the same or opposite sex of the participant. The age and race of the victim were controlled, all victims in the scenario were Caucasian and appeared to be approximately college age. The culpability of the victim was manipulated by having the victim behaviour in a culpable manner – walking away from the ATM machine while openly counting the money – or a non-culpable manner – putting the money in their pocket immediately after the ATM withdrawal. The results were consistent with the hypothesis that cognitive distortion serves to reduce threat by providing grounds for blaming an otherwise non-culpable victim.

Although there are numerous strengths of Marsh and Greenberg’s 2006 study, due to the special constraints, only two significant strengths will be discussed. First, and perhaps foremost, the study is the first to demonstrate capacity of motivational factors to distort memory for specific details of a criminal incident. In contrast, prior research within this field has focused on global distortions, including the general attribution of responsibility to a victim. By being the first to demonstrate this, the study serves as a platform for development and exploration of the personal motivations in an eyewitness account. Furthermore, another strength of this study pertains to its results. The results derive the same information as prior studies: faulty information processing occasioned by stress impedes accuracy of eyewitness accounts. However, the results diverge from the past by suggesting that such errors may be reflective of a defensive motivation – a desire to minimise threats to self. This would hold particular significance in court jurisdiction as it would question the overarching bias of eyewitness testimony, and, would potentially discredit it as a reliable form of evidence.

Of course, as is with any study, there are limitations. For instance, the participants were solely undergraduates, which could potentially impede the external validity of the results. The sample size, at such a specific type, makes the study less desirable for generalisation. Therefore, in order to remedy this, the study should have allowed for replicates with a diverse sample. Moreover, the participants were asked to recall information in a non-stressful environment, immediately after viewing. This decreases the suitability of the study for real life application as – due to large wait times surrounding court dates - witnesses may be asked to recall information after an extended period of time. In real life conditions witnesses are most likely to be questioned face-to-face under the stress and pressure of being in a courtroom. As such, real life relevancy could be inhibited.

Although justifiable, the factors concerning external validity are not of concern to the overall thesis. For the purposes of this essay, the study reinforces the unreliability of human memory in relation to recollection of past events. This is elucidated further considering that under more realistic conditions, the study’s results would most probably be magnified, with the added stress creating an expectation of amplified distortion.

Analysis of the Second Study

In addition to the previous study, research conducted by Morgan et al. assesses whether human memory for recently experienced, personally relevant, high stress events would be altered by exposure to suggestive misinformation. Participants were 861 – 649 male, 192 females, with an average age of 26 – active duty military personnel recruited for participation. The study hypothesised that exposure to misinformation - at either an individual or group level respectively – would results in falsified memories. The independent variable is the exposure to misinformation. The dependent variable, then, is the impact of misinformation on the accuracy of memory recollection.

The experiment engaged in four randomly allocated groups, all of which were run under high-stress conditions of Survival School training:

1. The Control Group, where participants were NOT exposed to misinformation during the training or questionnaire;
2. Misinfo-Questionnaire Group, where exposure to post-event misinformation was embedded in the questionnaire;
3. Misinfo-Photo group, in which exposure was within photographic misinformation;
4. Misinfo-Video group, where exposure was within a videotape concerning a specific event experienced as a group.

Group 4 was further divided into three sub-groups, of which were exposed to differentiating versions of the misinformation-videotape. All participant results were recorded via questionnaire – the groups vary based only on the time that the misinformation was received. The results confirmed the previously researched data showing that human memory for realistic, recently experienced stressful events is subject to substantial error. Furthermore, the results support the hypothesis that memories for stressful events are exceedingly vulnerable to modification by misinformation exposure.

There are clear, real-world strengths to the study conducted by Morgan et al. There is immediate concern regarding the vulnerability of human memory. Considering the high possibility that misinformation is being provided via social media and journalism alike, it could be prudent to begin to explore ways to reduce this impact, particularly in professional fields. Therefore, akin to Marsh and Greenberg, this study provides compelling evidence that practices such as eyewitness interviewing methods are not reliable. Particularly, that photographs and evidence viewed by a witness may induce false memory. This is elucidated by the experiment being conducted under high-stress, potentially further than that of a regular witness. This ensures the relevancy of the study in real world application. Furthermore, by utilising four separate conditioning methods for exposure, the experiment increases its reliability through replicates. Specifically, this displays that the brain can be manipulated to produce false memories in a multitude of circumstances.

There are, however, two limitations that should be discussed prior to evaluation. First, the level of stress induced in the experiment is exaggeratory compared to every-day, and, subsequently, may not be sufficiently generalised to the public. This is furthered through the allocated sample. Although the history of participants was presumably different, they were all active duty marine or navy personnel. Thus, the sample can be assumed to be relatively homogeneous. That said, military personnel should, theoretically, be more resistant to exploitation and propaganda efforts. Therefore, the memory of the general public can be induced to be more vulnerable. The experiment is also limited and cannot be expanded. This is due to the impact of the variating misinformation techniques was unable to be assessed. Subsequently, it remains unknown whether vulnerability to one technique indicates an increased risk of vulnerability for other types of misinformation.

Despite these limitations, the study maintains its importance and is directly reflective of the thesis, in that, memory should not be taken as an authoritative representation of past events. Instead, it is a malleable object, vulnerable to manipulation, and is, subsequently, unreliable.

In summary, it was argued that human memory, in relation to the details of past events, is unreliable. Two studies – Marsh and Greenberg and Morgan et al. - were used in support of this. The studies investigated the reliability of memory, focusing on bias and misinformation respectively, both results confirming the thesis. In addition to previously discussed remedies, future study is recommended to expand and develop the current findings.