Atkinson and Shiffrin (1968)

Foundation for Memory Research

Models of Memory: they are primitive diagrams of human memory to help understand the flow
of information and how it is stored.

Model 1: The multi-store model of memory (Atkinson and Shiffrin, 1968)

 The multi store model of memory is an explanation of the flow of information through a
series of stages.
 Atkinson and Shiffrin (1968) believed that memory involved three different stores. These
being sensory memory, short term memory and long term memory.
 They believed that to transfer information into the short-term memory it needs to be
attended to (attention is focused on the information in sensory memory) and to transfer
information from short-term memory into long-term memory the information needs to be
rehearsed.
 Atkinson and Shiffrin found that when information passes through these stores it can be
lost by forgetting.
 Their findings were that in sensory memory information is lost through decay, in short-
term memory information is lost through displacement and in long-term memory
information is lost as an affect of interference.
 Atkinson and Shiffrin also believed that rehearsal is needed for information to be
transferred to long-term memory and how well it is rehearsed will determine how
effective recall is.

STRENGTHS:

 New approach to memory, (people are information processors)

 Conceptualisation of memory as = support by research
 Possible to make predictions + design experiments based on model
 Has been modified e.g. by Baddeley and Hitch (1974) - working memory model

LIMITATIONS:

 Very simplistic, in particular when it suggests that both short-term and long-term
memory each operate in a single, uniform fashion.
 Can't account for how interaction between different stores happens
  Research on encoding of LTM has been challenged - now accepted that LTM contains
multiple stores (semantic, episodic and procedural)

Murdock (1962) – Serial position effect

Bennet Murdock (1962) studied the serial position effect in relation to short-term
memory, and his research supports parts of the multi-store model. Specifically, serial
position effects suggest both that STM has a limited capacity and duration, and that
rehearsal is necessary to encode memory from STM to LTM. Murdock's study focused in
on the serial position curve for free recall (Murdock, 1962):

 The aim of his study was to investigate how the serial positive curve was affected
by list duration and presentation rate.
 The sample included 103 participants, all university students, both male and
female, who were fulfilling a course requirement in introductory psychology. The
study was conducted in groups of five participants at a time.
 The procedure involved having different groups of participants listen to lists of
words with varying list lengths and presentation rates (lists ranged from 10-40
words in length, and the presentation rate was either one or two seconds). The
words themselves were selected from a list of the most common English words,
and Murdock had 80 different lists for each length/rate combination (participants
did 20 lists per session over four sessions).
 After each list, participants were given 90 seconds to recall as many words as they
could, in any order.
 The findings indicated that participants remembered more words from the
beginning of the list and the end of the list, which is known as the serial position
effect.
 Figure 3. Serial position
curves for the six experimental groups, showing list length and presentation rate.
Source: Murdock 1962
The serial position effect itself is shaped by two more specific factors: the primacy
effect (which refers to memory for items at the beginning of a list) and the recency
effect (which refers to memory for items at the end of a list). In theory, the recency
effect happens in Murdock's study because those words were still cycling through STM
(with its limited capacity of 7 +/- 2 items) at the time of the recall task. By contrast, the
primacy effect occurred because items at the beginning of the list had been rehearsed in
STM, and then stored in LTM, from which they were accessed during the recall task.
Presumably, those items at the beginning of the list had been displaced from STM by
incoming words in the middle of the list, but they had been rehearsed and encoded in
LTM, so they were later recalled.

Meanwhile, the items in the middle of the list were forgotten because they had already
been displaced from STM but had not been rehearsed for encoding in LTM. All added
up, the primacy and recency effects, as demonstrated by the serial position curve,
support the multi-store model's suggestion that STM has a limited capacity and duration,
and that rehearsal is necessary to encode memories in LTM.

Speaking more generally, Murdock framed the serial position curve as leading to
a significantly higher probability of recall for items near the end of a list (because of
STM) and slightly higher recall for items near the beginning of a list (because of LTM):
Glanzer & Cunitz – Primacy and recency experiment (Multi-
store model)
[A] Test primacy-recency effect.
 Participants were asked to read a series of 20 words.
[P]  They were then asked to recall the 20 words in any order.
 In another variation, a distraction task was performed before recall.
 Participants remembered the first and last few words better.
[F]
 Results reliably fall into a pattern known as the “serial position curve”.
 First few words – because they had more time to rehearse the words, encoding them into their long term
memory store.
[C]  Last few words – because it is still in the short term memory store.
 In the variation, the last few words were not recalled because of loss through decay.
 Provides evidence for multi-store model of memory.
 Low in ecological validity, lab environment.
 Ignored participant’s understanding of the words.
[E]
 Only one culture tested
 Education in some cultures may train students to remember things.

The results of this first experiment supported Atkinson and Shiffrin's multi-store model,
as it evidences the existence of multiple stores of memory.

Corkin (1997) – The case of HM

The case of HM (discussed previously in 2.A.3) is also related to the multi-store model,
in that HM's memory problems support the functioning of STM and LTM as independent
mechanisms operating in sequence. The basic idea is that HM's anterograde
amnesia evidenced severe malfunctioning of STM, and that in turn meant that any new
memories could not be encoded into LTM, just as the multi-store model suggests.

This was further supported by the case of Clive Wearing, who experienced many of the
same problems as HM. Furthermore, the fact that both men retained some procedural
memories and a few episodic memories (especially those with some emotional
connection) supports the idea that memory exists in separate cognitive stores. Basically,
both cases suggests that amnesia results from the breakdown of the encoding sequence
from STM to LTM, which is exactly what the multi-store model predicts.
Shallice and Warrington

Aim?

to illustrate a case where long-term memory remained intact when short-term memory
was damaged.

Method?

Case study of KF, who suffered a motorcycle accident causing damage to his left
parietal occipital region of the brain.

Results

KF showed very poor memory for numbers (usually less than two), but good
performance on tasks that seemed to indicate an intact long-term memory
Conclusion

Transfer of knowledge to long-term memory is not a linear process that always requires
an intact short-term memory. This therefore undermines the multi-store memory model.

Strengths
•Case study method allowed for in depth study of KF's memory issues.
•Findings seem to contradict the multi-store model of memory.
•Triangulation in this case study enabled the researchers to collect rich, deep data

Limitations
•Case study, therefore results cannot be directly generalised to broader population to
some extent.

Connection to WM
This study shows that short-term memory is not a static store, but a complex and active
information processor, composed of several dynamic subsystems, which is what WM
proposes.
Critical thinking
At this point, I ask them "so what?" How could they make this information about serial position effect
relevant in their own lives? Why is it important to know that in a list of information, we are most likely
to remember what is at the beginning and at the end of a list?
Here are some of the ideas my students come up with:

 When studying, we should not just read our notes from start to finish.
 In our school's speech contest, it is best to be first or last. But if you are in the middle, you need
to be an "artichoke."
 When remembering a movie, we tend to remember the ending - and not a lot of the details of the
plot.
 First impressions are important.