Priyanka IGCSE

Topic 2 – Memory
Atkinson And Shiffrin (1968) Multi-store Model of Memory

Atkinson and Shiffrin distinguished between the permanent structural features (cannot be altered)
of memory and its control processes (can be altered by how we encode, rehearse and retrieve
memories).

Structure of human memory – how we process information

Memory consists of three basic stores: a sensory register, short-term store and long-term store.

A sensory experience is obtained (though our 5 senses) from the environment. To store this new
information, we must first transform it into a form capable of being entered into the memory
system. This process of transforming sensory input into a memory trace is called encoding.

There are three ways of encoding information:

• Acoustic encoding – holding sound information.

• Visual encoding – holding images.

• Semantic encoding – holding the meaning of information.

• Elaborative encoding – relating new information to existing knowledge, ex: revising for an
exam.

Sensory Register/ memory.

This sensory input first enters the sensory register/memory.

There is a sensory register for each of these sensory modalities (seeing, hearing, touching, tasting
and smelling), making it to be known as modal specific. The sensory register can hold only a
limited amount of information for a few hundred milliseconds before it is lost.

However, if the information is paid attention to, then it is transferred into the STM.

Short-term store

Attended information enters the short-term memory, or the working memory, and is held
temporarily for 15-30 seconds. The STM allows information storage of about 5 to 9 items.

The encoding of information in the STM is primarily acoustic (auditory or verbal). This is because of
the phonological similarity effect.

Information is rehearsed through maintenance rehearsal to hold the information in the STM longer.
Elaborative rehearsal allows to transfer information from the STM into the LTM (A mnemonic can be
used to increase the strength of this transfer).
Priyanka IGCSE

Long-term store

Encoding into the LTM depends on the rehearsal process or some form of association between the
new and pre-existing knowledge stored there. The encoding of information in the LTM is largely
semantic.

The duration of the LTM is potentially a lifetime and a capacity which is potentially infinite.

When retrieving information from the LTM, we may experience the ‘tip-of-the-tongue’
phenomenon.

Retrieval (OUTPUT)

If information is neither accessible nor available then it cannot be retrieved from memory.

Retrieval may occur through:

o Cued recall: it is only possible to remember something when there is a prompt (cue) to
remind you.
o Recognition: information is presented and you can identify what you recall from what is
there in front of you.
o Free recall: you can retrieve information without cues, from ‘the top of your head’.

*Summary of the memory stores:

*Diagram to represent how we process information:

Priyanka IGCSE

Evaluation of MSM
Strengths
• There is strong supporting evidence for the model. Ex: Glanzer & Cunitz (1966) - a lab
experiment that demonstrated that STM and LTM may be separate memory stores AND The
case of Clive Wearing, who could form no new memories but had some intact LTM.
• The model is clear and straightforward, highlighting how memories go from environmental
stimuli to retrieved information.

Weaknesses

• The model is overly simplistic with its linear design and lack of detail per memory type.

• Lab experiments lack ecological validity and case studies of single participants cannot be
generalised so research into the MSM has only limited usefulness.

Evidence which supports the individuality of STM and LTM

Serial Position Effect
Evidence for STM and LTM being separate memory stores are supported by the idea of serial
position effect or primary-recency effect.

Murray Glanzer and Anita Cunitz (1966) conducted an experiment to investigate whether the
position of a word in a list affected recall. Participants recalled more word from the beginning
(primary effect) and end (recency effect), but recalled few from the middle of the list.

This was because words recalled in the beginning of the list were able to be rehearsed and
transferred to the LTM, while the words in the middle of the list were filling up slots in the STM. Then
the words at the end of the list would’ve displaced the memory trace of the middle words, and
instead now remain in the STM.

Coding
Priyanka IGCSE

The difference in the type of coding used by STM and LTM indicates we have separate memory
stores. Semantically related words were more difficult to recall from the LTM than acoustically-
related words.

Studies of patients with brain damage

Henry Molaison suffered from Amnesia following brain surgery for epilepsy, resulting in severe
impairment to his LTM, but his STM was largely intact.
Clive Wearing suffered LTM impairment following encephalitis, but his STM remained unaffected.
In both cases, patients were unable to transfer information from the STM to the LTM.
Clive Wearing could not recall past events in his life, but he could remember how to play the piano
and conduct an orchestra. KC (Kent Cochrane) could recall facts but showed severe memory
impairment in remembering personal events in his life before the accident. This suggests that we
have different LTM stores for storing specific memories (factual, skills and abilities,
autobiographical events).

Theories of forgetting

Cue-dependent forgetting

When we encode a new memory, we also store information such as about how we felt about it and
the place where it happened. When we cannot remember this information, it is because we are not
experiencing a similar situation now as we did before. Tulving’s encoding specificity principle
explains this as: ‘the greater the similarity between encoding and retrieval event, the greater the
likelihood of recalling the event’.

Context-dependent forgetting

If we are not in the same situation as when learning, we may not be able to access the information
easily. Location and other environmental cues can aid recall or impair if it is missing.

State-dependent forgetting
Priyanka IGCSE

When we learn information, we encode details about our emotional and physical state at that time.
Being in the same state of mind during encoding and recall helps retrieval of information to be more
successful.

Trace decay

The theory suggests that learning causes a physical change in the neural network of the memory
system, creating a memory trace or an ‘engram’. Once this memory trace has been created, it must
be reinforced through repetition and practice; if not, it will weaken and decay, causing forgetting to
occur.

Interference

In the LTM, forgetting can occur due to interference or confusion between old and new memories.
This means that the memory does not get lost but may become distorted or changed as a result of
conflicting memories. It happens through retroactive and proactive interference.

Displacement

As new information enters the STM, it may overwhelm the initial, temporarily stored items in it and
cause them to be ‘pushed out’ – displaced by the new.

Individual Differences in memory

The speed at which we process information differs between individuals. This is known as
processing speed. Processing speed and STM capacity is affected by age as well. Younger children
have a shorter digit-span than older children which suggests that their memory capacity increases
with age.

Amnesia

Amnesia is a form of memory loss as a result of brain injury, genetic factors, ect. Amnesia affects
the LTM.

Anterograde Amnesia

The individual is unable to form new memories after the onset of Amnesia. (ex-: a person diagnosed
with Anterograde Amnesia, after an accident, can remember events from before the accident.
However, is unable to form new memories after the accident.

Retrograde Amnesia

The individual loses memory of events that took place before the onset of Amnesia. (ex-: a person
diagnosed with retrograde amnesia, after an accident, can make new memories but is unable to
recall any memories before the accident.

Working Memory Model (Bradley and Hitch, 1974)

Priyanka IGCSE

Alan Braddley and Graham Hitch proposed the STM (working memory) as consisting of three
components: a central executive that would deal with the running of the system, and two slave
systems to deal with visual and verbal information.

The role of the working memory would be to temporarily store and manipulate information being
used. The working memory is fragile, so it is susceptible to distractions, overload, and overwork.

Reconstructive memory (Bartlett, 1932)

Our memories do not replay the exact version as how we remembered it,
instead we only have fragments of the original event stored. As a result, Bartlett
viewed memory as constructive in nature.

We fill in the gaps of missing information through what we think suits it and
using our past experience (confabulation) to make it make sense – this
process of active recall is reconstructive memory.

Schema Theory

Schemas are mental representations of information we have about a specific object or event. Every
schema has fixed and variable information. How we interpret and reconstruct these schemas when
we are remembering them is referred to as Active Reconstruction. Reconstruction takes place by:

1. Omission
2. Familiarization } leveling- involves removing or downplaying details from the memory

3. Rationalization
4. Transformation } sharpening- involves adding or exaggerating details to the memory.

1) Omission

Is where information is left out completely as the individual does not have schemas available to
understand the information present.

2) Familiarization

Is where information which seems unfamiliar is altered in such a manner where it fits with currently
existing schemas.

3) Rationalization

Is where information is added so that a new schema could be made which would make sense to the
person.

4) Transformation
Priyanka IGCSE

Is where information is changed as a whole, which means that transformation is a product of

familiarization, rationalization and occasionally omission.

Does this make eyewitness unreliable? - Loftus argued that Reconstructive Memory implies
that eyewitnesses to crimes will often be unreliable.

Bartlett (1932) War of Ghosts Study

Aim: To investigate the effect of (cultural) schemas on reconstructive memory

Sample: 20 male undergraduate students from the University of Cambridge in the UK

Procedure:

1. Bartlett used serial reproduction:

The first participant read the Native American folk story, The War of the Ghosts which he/she then
reproduced in writing from what they could remember of the story.

This written recall was then read to a second participant who then wrote their own memory of the
story, which was then read to a third participant who too produced their own version of the story
and so on.

Results: The participants’ stories bore little similarity to the original Native American folk tale. The
changes made by the participants included:

Omission: Key details of the story were ignored or missed out:

• unfamiliar or unpleasant details such as a contorted face or black coming out of a mouth
were ignored.
• ghosts as key figures in the story were soon dropped from the re-telling of it.

Assimilation and sharpening: Story details were changed to suit the participants’ own cultural
schemas:

• ‘canoes’ became ‘boats’

• ‘paddling’ became ‘rowing’
• details such as a ‘spirit wound’ were re-interpreted as a flesh wound

Levelling: The story became shorter - the original story was approximately 350 words and the
participants’ version was around 180 words.

Conclusion: Cultural schemas contribute to the reconstructive nature of memory. Memory is an

active process in which pre-existing information and expectations may interfere with the accuracy
and reliability of memory.
Priyanka IGCSE

2. Bartlett asked participants to read and recall a North American folk tale called ‘The war of
ghosts’.

This folk tale was chosen because:

I. It was culturally unfamiliar to participants (allows to examine the transformations when the
story is reproduced).
II. It lacked any rational story order.
III. Dramatic nature to the story (would encourage visual imaging).
IV. Conclusion was somewhat supernatural (how would they perceive and image this?).

Each participant read the story twice and repeated reproduction was used to test the effect of
time lapse on recall. Twenty participants recalled the story after several minutes, weeks, months
and years (longest time lapse was six and a half years). Bartlett found out that the story became
considerably shortened (because omission were made), phrases used reflected modern concepts
and story became more coherent in form.

Bartlett concluded that memory is reconstructed each time it is recalled.

Evaluation:
Strengths of the study
• Remembering a story is more naturalistic than nonsense trigrams and therefore gives the
study ecological validity.
• The participants were unfamiliar with the story, therefore decreasing bias of findings.
• There is evidence from other studies (Loftus and Palmer (1974)), so this increases reliability.
• Data was gathered through qualitative analysis so the real nature of reconstructive memory
can be understood.

Weaknesses of the study

• Bartlett was criticized for his use of the ‘War of Ghosts’ story for having little relevance to
everyday memory. However, Bartlett conducted his repeated reproduction experiments on
eight different stories and concluded the same results.
• The story was not familiar and was illogical so remembering the story is not a mundane
task.
• Using qualitative analysis to gather information is unscientific.
• A standardized procedure was not followed since participants were made to recall the study
at different time intervals and the story was allowed to be read at participant’s normal
reading pace, therefore making the study unscientific.
• The study's report contains limited and missing details of how the experiment was carried
out, making it difficult to replicate the study to measure its reliability.
• The study was conducted on a limited sample of 20 people of only UK college students, and
therefore, the study is non-representative and likely to be non-generalisable.
Priyanka IGCSE

• It is unclear as to whether the memory distortions were the product of schematic

interference or individual differences such as: poor memory generally, lack of attention.

Peterson and Peterson (1959) Short-term retention of individual verbal items

Aim: to test the true duration of the STM.

to provide evidence for the MSM.

Sample: 24 Psychology students.

Procedure:

The experiment took place in a lab setting where each participant was tested individually. After
having been recited by the experimenter, each student had to recall meaningless three-letter
trigrams at different intervals (delayed by 3, 6, 9, 12,15 or 18 seconds).

(In the first experiment) To prevent rehearsal (practice), the students had to count backwards in
threes or fours from a specific number until they were asked to recall the letters. This is known as
the Brown Peterson technique.

(In the second experiment) Participants were asked to do the same tasks but some were given time
to repeat the trigram before counting backwards.

Results:

In the first experiment, the longer each student had to count backwards (the distraction task), the
less they were able to recall the trigram accurately.

• 80% of the trigrams were recalled correctly after 3 seconds.

• This fell to 50% after 6 seconds.
• Less than 10% were recalled accurately after the 18-second delay.

This proves that STM has a limited duration when rehearsal is presented.

In the second experiment, extra time increased the frequency of recall (they were able to
consolidate the information a little more). Although, there was a similar decline overtime.

Conclusion:

Peterson & Peterson concluded that STM has a limited duration of approximately 18 seconds.
Furthermore, the results show that if we are unable to rehearse information, it will not be passed to
long-term memory.

Evaluation:
Strengths of the study
Priyanka IGCSE

• Controlled, standardized experiment – study was carried out in a lab setting, the
researchers used fixed timings, they eliminated noise/other factors that could influence
memory. So, the study can be replicated with reliability.
• By using the distraction task to see how it affects the ability to retain information, it has high
internal validity to ensure the study is measuring what is intended.
• Real life application – suggests that we should learn in bits and rehearse them to retain
information.
• Provides good evidence for MSM.

Weaknesses of the study

Research uses nonsense trigrams because real words could have personal relevance, but it tends
to lack mundane realism and ecological validity.

It lacks generalizability:

• All participants were psychology students

• Memory of psychology students may be different from other people.
• They may already have encountered MSM previously.

Reductionism Vs Holism

Reductionism Holism
Considers human behaviour as a whole rather Aims to break down psychological processes
than as parts. into small and simple elements.
Based on qualitative analysis (own Based on scientific assumption.
interpretations).
Includes humanistic, social and Includes biological, cognitive and behavioral
psychodynamic approaches. approaches.
Considers the context and complexity of Seeks to isolate and study individual
human behaviour. components in isolation.
Ex – Reconstructive memory. Ex – Multi-store Model of Memory.