To what extent is one cognitive process reliable? [22]
Memory is a cognitive process that many psychological studies have evaluated for reliability. Memory is
the storage and subsequent retrieval of information stored within the brain; it is concerned with how we
process information in the present based on what we already know. Through research, it has been shown
that memory is reconstructive in nature and hence we do not store complete replications of events, but
rather our memories are subject to bias based by the way we perceive information during encoding, and
by how we process the memory of event in accordance with pre-stored schemas. Reconstruction of
memory can hence be influenced by many factors, including post-event information, sociocultural schemas
and the emotional significance of the event. This essay will attempt to determine to what extent the
cognitive process memory is reliable, with regards to its reconstructive nature.
Schema theory, first explained by Bartlett in 1932, attempts to explain the reconstructive nature of
memory. A schema is a mental representation of knowledge stored in the brain; it can be seen as a
network of knowledge, beliefs and expectations about particular aspects of the world, which is built up
through experience with the world and is subject to change. Schemas serve as cognitive shortcuts; they are
used to organize our knowledge through categorization, assist memory recall and guide our behaviour.
They simplify reality through setting up expectations about what is probable in relation to particular
environments, and allow us to make predictions about certain occurrences and makes sense of our current
experiences. Schema theory argues that the information we store and retrieve in our memory is subject to
alteration based on the information we have already stored in our schemas, and hence is constantly
subject to change, especially if it is uncertain or unambiguous. Humans actively seek meaning to the world,
and in the absence of certainty, schemas can be utilized.
Bartlett aimed to investigate how cultural schemas affected the reliability of memory in his seminal study
‘The War of the Ghosts’ [1932]. Bartlett gathered 20 British students and asked them to listen to a story
called ‘The War of the Ghosts’, which originated from a North American Legend and used vocabulary and
sentence structure unfamiliar to British language. As the content and development of the story was
unfamiliar, this meant that the participants were more likely to rely on schema processing to understand it,
by relating it to their own experiences. He then asked the participants to reproduce the story after a short
time, and then over a period of months or years (serial reproduction). Bartlett found that the participants
changed the story as they tried to remember it in a process called distortion; there were three patterns of
distortion that took place. The story was assimilated, meaning that the content and development of it
became more consistent with the participants’ own British cultural expectations. The story was levelled,
meaning that with each subsequent reproduction the story omitted more information that was seen as less
important. Finally, the story was sharpened in response to it’s unfamiliar structure and development,
meaning that the order of events was rearranged in order to make sense of it from the participant’s own
cultural perspective. The participants retained the main themes of the story but changed the unfamiliar
elements to match their own cultural expectations so that the story remained coherent. From these
results, Bartlett concluded that memory is an active process, where information is retrieved and altered to
fit into existing schema, in order to make sense of unfamiliarity. Therefore, memories are not replications
of information, but rather they are reconstructions, combining original information with information
retrieved from schemas to create an overall impression of an event or concept. Therefore this study
supports the argument that memory is inherently unreliable.
There are some issues with the generalizability of the results of Bartlett’s study however. Firstly, only
twenty participants were used in the experiment so it has a very small participant sample, and all of these
participants were undergraduates of Cambridge University in the UK during the 1930’s. Therefore it is likely
that all participants were from very similar social and cultural backgrounds, meaning that there is a lack of
participant variability. As all these participants were all of Western culture and had similar social
environments through their enrolment in higher education, this study only tested the effects of cultural
schemas on one specific culture of people, and therefore the findings cannot be generalised to other
, cultures and lack population validity. Also, it is invalid to conclude that it was cultural schemas that caused
the memory distortion, as there was no control group of North American participants with whom the
results could be compared. If the story were distorted in a similar way in both cultures, it would be due to
schema processing in general rather than cultural specificities. The validity of the results is also questioned
by the lack of control that Bartlett had over the time between reproductions, and the contact between
participants during the time period. These issues with validity mean that the conclusion of the study in
arguing that memory is reconstructed via schematic processing may not be correct, however Bartlett’s
research was significant in stimulating further research within the field of reconstructive memory which
confirmed his results.
Another study that aimed to investigate how schemas affect the reliability of memory was conducted by
Loftus and Palmer in 1974. Loftus hypothesized that the memory recall of eyewitnesses could be
influenced by post-event information in the form of leading questions. A leading question serves to
encourage a particular response through exploiting schematic associations with certain words. In the
study, 45 students were shown videos of short car accidents and were then split into 5 groups of 9. Each
group was asked a critical leading question, but the active verb used within the question operated as the
independent variable: ‘How fast were the cars going when they contacted/ bumped/ hit/ collided/
smashed into each other?’. The study found that as the verb became more suggestive of speed, the mean
speed estimate of the group increased. ‘Smashed’ yielded he highest mean at 40.8 mph, and ‘contacted’
led to a mean of just 31.8 mph. The verb used in each condition was associated with a different schema
that gave extra information about the severity of the crash and how far the cars were travelling. This study
supports Bartlett’s theory that memory is reconstructive in nature and is therefore unreliable, as it can be
effected by post-event information.
A large issue with this study is that is cannot determine whether the participants’ memories of the event
were actually altered by the schemas triggered through the verbs in the leading question, or whether the
verbs just biased the participants towards a particular answer. If it is the latter, then it cannot be claimed
that memory is reconstructive in nature as Schema Theory claims. Either way however, the study suggests
that memory is unreliable. Some issues with this study arise from the experimental procedure. The
participants only watched the car-crash in video format, meaning that the event lacks the emotional
impact it would have in reality. This means that the participants may have acted differently to how they
would have done if it had been a real occurrence, and therefore the results of the study lack ecological
validity. This means that the results cannot be accurately used to predict the reliability of memory when an
event does have emotional impact, for example eyewitness testimonies are usually related to emotionally
poignant events.
A Theory that links considers how the emotional impact of an event affects its reliability is Flashbulb
Memory Theory. Flashbulb memory is a special kind of emotional memory, which refers to vivid and
detailed memories that appear to be recorded in the brain with a high degree of accuracy, like a photo. It is
assumed that these memories remain intact and accurate due to the emotional arousal during encoding.
Brown and Kulik suggested that the memories may remain intact and accessible due to being rehearsed by
the individual, due to its emotional significance. There appears to be empirical scientific evidence for the
role of emotion in memory, as modern neuroscience suggests that emotional memories are better
remembered due to the critical role of the amygdala, which plays a role in emotional processing.
A study by Yuille and Cutshall [1986], like Loftus and Palmer, aimed to investigate the role of leading
questions in the accuracy of memory recall. They believed that the results of Loftus and Palmer’s study was
inaccurate due to a lack of ecological validity, and hence the participants of their study were eyewitnesses
to a real theft and gun-shooting crime. The police originally interviewed all the witnesses and 20 of these
eyewitnesses were contacted 4-5 months after the event to take part in the study (13 of them agreed). The
experimenters asked questions in the same way as the police did; allowing the participants to give an
account of the event and then asking questions. Two leading questions were used in order to test the