The MSM is a structural model
of memory developed by
Atkinson and Shiffrin, with an
emphasis on the rehearsal of
information. It suggests that
the STM and LTM are two
unitary, singular stores.
Information passes through
each store in a linear process.
Each store has different
encoding, durations, and
Information from our
environment initially goes into
the sensory register where it
stays for about a second. We do
not notice much of this stuff, but
if we pay attention to it, the
information will pass into our
short term memory. Information
can be kept here by maintenance
rehearsal, but if it is further
processed through elaborate
rehearsal, it moves into our long
term memory. The information
there can then last forever.
Brain damaged patients- HM had
a part of his brain removed to
stop his severe epilepsy. As a
result, he suffered from
anterograde amnesia. He could
not form any new long-term
memories. This showed that the
STM and LTM are separate stores
and that they can work
Serial position curve- Murdock devised the
serial position curve explaining the primacy
and recency effect, Experiments show that
when participants are presented with a list of
words, they tend to remember the first few
and last few words and are more likely to
forget those in the middle of the list. This
gave evidence for the LTM and STM stores,
murdock was based his work on the MSM.
The MSM is said to be oversimplistic. It
places too much emphasis on the role of
rehearsal and does not account for
situations where we learn information
instantly without rehearsing it. It is likely
that semantic coding takes place in these
circumstances. Additionally, research
suggests that the LTM is not a unitary
store, we have one store for semantic
memories (factual) and a different one for
procedural ones (actions). The MSM does
not reflect this.
The MSM does not account for more than
one STM store. The case study of KF, an
amnesia patient, showed that he had
difficulty recalling digits read to him, but
was much better at recalling when he read
them. This suggests there is separate
stores for visual and auditory information,
which is not covered by MSM.
Working Memory Model
Developed by Baddeley and Hitch, the
WMM replaces the idea of the unitary
STM store, bases on rehearsal, as
proposed by the MSM. It is an active
system where mental tasks are carried
out on a conscious level rather than an
unconscious one, it emphasises the
importance of processing. Each
component has a limited capacity.
The Central Executive-
Modality free; most
It operates all the other
slave systems and
Its capacity is very
Pad- Used to store
and manipulate visual
patterns and spatial
arithmetic tasks and
helps the individual
The Phonological Loop-
Deals with auditory stimuli
where information needs
to be rehearsed to prevent
decay. Contains the
Phonological store; 'inner
ear' (sound speech) and the
process; 'inner voice'
The WMM has research support. The case study of KF, an amnesia patient, showed that
he had difficulty recalling digits read to him, but was much better at recalling when he
read them. This suggests there is separate stores for visual and auditory information,
which is showed by the WWM. His phonological loop would have been damaged.
-The central executive is unfalsifiable; Richardson criticises the vague
description of the CE despite it's crucial role. -There may be further stores
which are not mentioned such as a musical component since we are able to
listen to non-lyrical music without disrupting other stores.
integrates the visual,
spatial and verbal
information from the
other stores together.
Links to LTM.
Baddeley found participants had difficulty carrying out two visual
tasks at the same time, but could do a visual and verbal task at the
same time. The greater difficulty is because both visual tasks compete
for the same resources (VSS). This supports the different VSS and PL
Theories of forgetting
Motivated forgetting where emotionally threatening and
traumatising events/memories are thought to be banished
from our LTM into the unconscious mind to avoid
re-experiencing the anxiety it may cause.
Williams investigated women who
had suffered childhood sexual
assaults. 38% had no recall of
events, 16% reported they hadn't
once been able to.
Interference- Forgetting occurs
when one memory blocks another,
causing one or both to become
distorted or forgotten.
Retroactive interference- where NEW information interferes
with the ability to recall OLDER information.
Proactive interference- where OLDER information interferes
with the ability to recall NEW information.
Underwood & Postman- supported retroactive interference; gave groups lists of
word pairs such as cat-tree. One group had to learn one list, one had to learn two.
Both had to recall the first. Recall was worse for the group that learnt two lists.
Most research is lab based, providing evidence for
both types of interference whilst keeping control of
extraneous variables to improve validity.
However, artificial materials are used. The material is often
word lists or groups of numbers, which is different to
day-to-day remembering. This may make interference way
more likely due to the individual not being interested in it.
Interference is worse when
memories are similar.
McGeoch and McDonald- Participants had
to learn a list of 10 words until they could
remember with 100% accuracy. They then
learned a new list, in 6 different groups all
learning different types of lists. 1:
synonyms. 2: Antonyms. 3: related words.
4: nonsense syllables. 5.three-digit
numbers. 6: no list, rest. Participants recall
depended on the nature of the second list-
similar material produced the worst recall
(synonyms). Shows interference is
strongest when memories are similar.
Baddeley and Hitch asked rugby players to recall names of teams they had played so far,
week by week. Accurate recall did not depend on how long ago the match was, but the
number of games played in the meantime. This applies to everyday situations.
Context dependent- Occurs with external
retrieval cues; the environment is different at
recall to when the material is learnt.
Godden & Baddeley- Divers learnt information either on
dry land or underwater, and then recalled it either on land
or underwater. Recall was found to be 40% worse when
the recall context was different to the environment in
which it was learnt, compared to when the environment
was kept the same. This demonstrates
context-dependent forgetting because information wasn't
retrievable when the environment did not match.
State dependent- Occurs with
internal retrieval cues; internal
environment is different at recall
than to when the material was
Darley- found participants who hid
money while high were less able to recall
where the money was when they were
then not high, compared to when they
were high again.
When information is placed in the memory,
associated cues are stored at the same time. If
these cues aren't available at the time of recall, you
may not be able to retrieve the information.
Baddeley argued that
different contexts have
to be very different
indeed before an effect is
seen (i.e land vs water).
Learning something in
one room and recalling it
in another is unlikely to
result in forgetting.
Therefore real life
applications may be
weak and don't actually
explain much forgetting.
Godden and Baddeley
replicated their underwater
experiment but using a
recognition test rather than
recall. There was no
Performance was the same
over all conditions. This limits
retrieval failure as an
explanation because it only
affects memory when testing
recall rather than recognition.
A strength is that there are real life applications. People often report going downstairs
to get an item but then forgot what they came downstairs for. They then go back
upstairs and remember again. This suggests that when we have trouble remembering
something, we should revisit the environment in which we experienced it. This can be
applied to the cognitive interview. Eyewitnesses are told to reinstate the context when
recalling events so that their memory is jogged by environmental cues.
Eye witness testimonies
Post event discussion
When co-witnesses discuss a crime, they
mix (mis)information from the other
witnesses with their own. Witnesses go
along with each other to win social
approval or because they believe the
other witnesses are right.
Gabbert et al. - Participants were asked to
watch videos of a crime, but filmed so each
participant saw different parts than the other.
Both then discussed what they saw afterwards
before completing a test of recall. 71% of
participants mistakenly recalled parts of the
video that they did not see but instead picked
up in the post-event discussion. In the control
group with no discussion, there was no errors.
Leading questions do two
things: The wording of the
question has no enduring
effect on an eyewitness'
memory of an event, but
influences the kind of answer
given, or the wording of the
question does affect eye
witness memory and
interferes with the original
memory, distorting its
Loftus and Palmer- 45 university students
watched film clips of car accidents and then
answered questions about speed. Critical
question: 'How fast was the car going when
it hit the other car?' The participants were
split into 5 groups, with the verb changing
between each; hit, contacted, bumped,
collided or smashed. Powerful words like
smashed gave on average higher speed
estimates (40.5) and softer ones such as
contacted gave on average lower speed
estimates (31.8). The leading verb biased the
eyewitness recall of the event.
Has good real-life applications. The research has led
to important practical uses for police officers and
investigators because the consequences of
inaccurate EWT can be very serious. This has led to
an improvement in the interviewing system of EWT.
Loftus and Palmer used artificial materials. Watching a film
clip of an accident it very different to witnessing a real
accident in real life. Yuille and Cutshall argued that anxiety
has an important role in remembering events. When
watching a clip, we are alert and focused on details, whereas
in real life, we are impacted by stress and fear which could
impact our memory. This makes application difficult.
Lacks external validity.
What an eyewitness can
remember in real life
situations can have
such as a successful
conviction. This is not
true for research studies,
where they may put less
effort into trying to
Participants want to
be helpful and
attentive. So when
asked a question they
don't know the answer
to, they guess. This
challenges the validity
of EWT research.
Anxiety has strong emotional
and physical effects. But it is
not clear whether these
effects make eyewitnesses'
recall better or worse.
curve explains a
a level of anxiety
accuracy of EWT,
but as anxiety
Anxiety has positive effect
Yuille and Cutshall- 13 witnesses of a
real life crime of a gun-shop owner in
Canada, were interviewed 4-5 months
after the incident and rated how
stressed they were when it happened.
Witnesses were very accurate and
there was little change after 5 months.
Those who experienced the highest
levels of anxiety were most accurate
(88% compared to 75%).
Field studies lack control of variables. Real life witnesses are
interviewed some point after the event. Many things happen to
them in the meantime that researchers cannot control. For
example, discussing with other eyewitnesses, read accounts or
see images through the media, etc. These variables may be
responsible for the accuracy of recall.
Demand characteristics- participants are aware they are
watching a film for a reason to do with a study. They may
work out that they will be asked questions about what they
have seen and so may pay extra attention to details or give
responses which they believe will be helpful.
Creating anxiety in
participants is potentially
unethical because it may
subject people to
Ethical issues don't
challenge the findings of
studies, but they do raise
conducting the research.
So real-life studies are
beneficial because the
participant has already
witnessed the event.
Anxiety has negative effect
Loftus explained the weapons focus effect,
whereby a person focuses on the weapon
involved when in a crime and therefore
ignores any other details.
Johnson & Scott- Participants sat in a waiting room believing they were
about to take part in a lab study. Participants heard an argument in the
next room; low-anxiety condition: participants then saw a man walk
through the waiting room carrying a pen with grease on his hands; high
anxiety condition: participants heard the sound of breaking glass and
then a man walked through the waiting room with a bloodied knife. 49%
of participants in the low anxiety condition could correctly identify the
man, 33% of participants could identify him in the high anxiety
condition. Participant's focus was on the source of danger and not the
man's face. Showing anxiety worsened their memory.
May have tested surprise not anxiety. Participants may focus on the
weapon because they are surprised by it rather than because they're
scared. Pickle used scissors, a handgun, a wallet and raw chicken as
hand-held items in a hairdressing salon. EWT accuracy was worse for
high unusualness (chicken and handgun). The weapon focus effect
may be due to unusualness rather than anxiety/threat.
Fisher and Geiselman claim that EWT could be
improved if the police use techniques based on
psychological insights of how the mind works. It is
based on cognitive psychology.
every detail of the
event, even if it
seems irrelevant or
the witness is not
confident about it.
Reinstate the context:
Witnesses return to the
scene 'in their mind' by
describing the environment
(weather, smells and
Reverse the order- Events are recalled in a different chronological order (e.g from
end to beginning or from the middle to beginning) this prevents them using their
expectations of how things would have happened.
Change perspective- witnesses recall the event from
other people's point of view and perspectives.
Holliday- children ages 4-5 and 9-10 watched a 5
min video of a child's birthday party, and were
interviewed the next day about it using either
SPI or CI. CI produced more accurate details
It's time consuming. Police are reluctant to
use CI over standard interviewing because
it takes up much more valuable time.
Additionally, it also requires special
training which is limited by time and
money. This makes it unlikely that a proper,
full version of CI is used.
Studies of the effectiveness of CI
inevitably use slightly different
researchers may use variations
of CI and involve their own
methods. This reduces our
ability to draw conclusions due
to poor reliability.
Types of Memory
Coding, capacity and duration
CAPACITY STM- 7+/-2
Miller- made everyday observations.
Found people remembered things in
sevens, noted 7 days of the week and 7
notes on a musical scale etc.
CAPACITY LTM- UNLIMITED
CODING STM- ACOUSTIC
Baddeley- Groups had to recall either
acoustically similar or dissimilar words or
semantically similar or dissimilar words.
Acoustically similar words recalled worst
instantly and semantically similar recalled
worst after 20 mins.
CODING LTM- SEMANTIC
DURATION STM- 18-30s.
Peterson and peterson- students were given a consonant
syllable (YCG) to remember and a 3 digit number to count
backwards from for either 3,6,9,12,15 or 18 seconds. 80%
were recalled after 3s. 3% recalled after 18s.
DURATION LTM- LIFETIME
Bahrick- Participants had to recall
names from their graduating class
using free recall and photo
recognition. 48 years later photo
recognition was still 70% accurate.
didn't use meaningful
info, people may use
semantic coding even
for STM tasks. It
therefore has limited
capacity was more
like 4 chunks.
Peterson and Peterson
is artificial. memorising
does not reflect real life
where we often
Bahrick has high
(people's faces and
Things we 'know that'.
Episodic: Gives an individual an autobiographical
account of personal experiences e.g birthdays.
Helps individuals distinguish between real life
events and imagined events. It is influenced by
emotion at the time the memory was encoded.
-The extent to which episodic and semantic memories are
different is unclear, as although different brain areas are
involved, there is a lot of overlap in which semantic
memories originate from episodic ones. Cohen and Squire
argued that there are two types of LTM- declarative and
non-declarative, meaning semantic and episodic memories
are stored together the same way, and procedural is a
distinctly different type, stored differently.
Tulving- had participants perform memory tasks whilst having a
PET scan. Episodic and semantic memories were both in the
prefrontal cortex but semantic was to the left, and episodic to the
right. This shows the physical reality in the different LTM types.
Semantic: contains all knowledge
(facts, concepts and meanings) a
person has learnt. It is less personal
and more about knowledge people
share. Often linked to episodic LTM's
as new knowledge tends to be learnt
-May involve more of a network
of associated links performed in
different brain areas, rather than
a single form of memory ability.
Some links may be stronger than
others explaining why some are
easier to recall than others..
Vicari et al. Case
study of CL- an eight
year old girl who
suffered from brain
damage due to the
removal of a
deficiencies in her
but was still able to
create and recall
Things we 'know how' to do
Procedural: Occur mainly in early life
involving important motor skills such as
talking, walking and dressing. It allows
individuals to carry out learnt tasks with
little to none thought - we remember
without conscious awareness.
-There is lack of case studies of people with brain damage that
affects procedural memory but not declarative memory, meaning
there is a lack of supporting evidence. - Procedural memories take
longer to learn than declarative memories. this may be due to the
involvement of motor functions and spatial abilities that declarative
memories tend not to include.
Finke et al. case study of PM, a 68 year old professional musician who
suffered from brain damage resulting in amnesia. His episodic and
semantic LTM were affected, his procedural was not effected and he
could still play musical instruments.