Transcript MEMORY - stcmpsy

MODELS OF MEMORY
DEFINITION OF MEMORY:
“The retention of learning or experience”
MEMORY









MEMORY ANALOGIES
Library
Aviary
Underground map
Computer
STAGES OF MEMORY
ENCODING
STORAGE
RETRIEVAL
MEMORY








THE SHORT TERM MEMORY
DEFINITION: “Memory for events in the present or
immediate past”
CAPACITY OF THE STM
The Digit Span Test (Jacobs 1887)
Magic number 7 (Miller 1956)
Note: Vogel and Cowan (2001) have concluded its
more likely to be 4 items
Chunking ( a useful application of psychological
theory)
rehearsal
THE SHORT TERM MEMORY
ENCODING IN THE STM
 Predominantly acoustic (sound)
 Evidence = errors made with similar
sounding letters when presented visually for
0.75 seconds (Conrad 1964)
 Note: research has shown we use other
forms of coding also e.g. visual
 DURATION OF THE STM
 Approx. 15 – 30 seconds without rehearsal

THE SHORT TERM MEMORY

A STUDY INTO THE DURATION OF THE
STM (Peterson & Peterson 1959)

See worksheet
THE LONG TERM MEMORY
DEFINITION: “Memory for events that have
happened in the past”
 CAPACITY OF THE LTM
 Unknown
 Unlimited

THE LONG TERM MEMORY






ENCODING IN THE LTM
Predominantly semantic
Evidence: (Baddeley 1966) semantically similar
words presented to the LTM are most susceptible
to being muddled up
But also acoustic, visual,
olfactory, gustatory
DURATION OF THE LTM
Potentially a lifetime but there
are many individual differences
THE MULTI-STORE MODEL OF MEMORY
(Atkinson & Shiffrin 1968)
DIAGRAM
New idea of sensory memory
Short term and long term memories are permanent, structural components
SUPPORTING EVIDENCE (mini experiment)
1.
PRIMACY/
RECENCY
EFFECT
(Murdock
1962)

The above diagram is called a serial position curve
and is produced when a word list is free recalled.
PRIMACY EFFECT =
RECENCY EFFECT =


2. DIFFERENT TYPES OF CODING IN STM & LTM
Baddeley (1966) studied STM
 P’s presented with acoustically similar words
 P’s then presented with semantically similar
words
 Results = immediate recall of acoustically
similar words was poor because the
grammatical and phonemic parts of language
are quickly forgotten

3. USE OF DIFFERENT AREAS OF THE BRAIN
•
Using brain scans Beardsley 1997 found that
people use the prefrontal cortex when doing a
STM task and the hippocampus when doing a
LTM task
4. STUDIES OF PEOPLE WITH MEMORY LOSS
See video and worksheet
EVALUATING THE MULTI-STORE
MEMORY
POSITIVES
 Has evidence to support it
 Looks at structure AND process
 Enables us to make predictions
EVALUATING THE MULTI-STORE
MEMORY
NEGATIVES
 Oversimplified
 Unitary stores
 Proposal of one short term store is wrong
(see next model)
 Proposal of one long term store is wrong
EVALUATING THE MULTI-STORE
MEMORY








Cohen & Squire (1980) distinguished between:
declarative memories (memories for ‘knowing that’) and
procedural memories (memories for ‘knowing how’) –
people with different conditions can lose one or the other,
e.g. HM his STM was damaged but he could still learn skills
(learned to play tennis)
Clive Wearing – musical ability still intact
Sub-divisions of declarative = semantic
episodic (Tulving 1972)
Evidence for this in brain scans using radioactive gold to
measure blood flow
Model finds it difficult to explain flashbulb memories (these
can be autobiographical or historical)
Examples are:
EVALUATING THE MULTI-STORE
MEMORY







Rehearsal – now accepted that is not to only way
into LTM
Levels of processing model (LOP) – depth of
processing is more important
3 levels i) visual ii) phonetic iii) semantic
Integrated STM and LTM not separate
STM relies on LTM’s to chunk for instance
e.g. AQABBCITVIBM
Supporting evidence comes from artificial, unecologically valid laboratory experiments
The Working Memory

See work sheet
Evaluating the Working Memory
Model
POSITIVES
 Can explain partial memory difficulties
 e.g. case sudy of someone with normal LTM
but phonological loop difficulties
 Has plenty of research evidence, e.g. dual
task experiments
 Emphasizes the active nature of short term
memories
Evaluating the Working Memory
Model


Ties in with brain
mapping technology
Brain imaging studies
have shown the
separate areas at
work, e.g.
phonological store in
Wernicke’s area and
the articulatory
rehearsal process in
Broca’s area
Evaluating the Working Memory
Model
NEGATIVES
 Little is known about the CE (vague concept) – its
probably also subdivided
 Using case studies of brain damaged people is
problematic
 Lab studies also have their drawbacks ????????
ACTIVITY
“Alice is ……

An alternative to the multistore model by
Craik and Lockhart (1972)
Emphasises memory process rather than
structure
 Based on the idea that the strength of a memory
trace is determined by how the original info was
processed

www.psychlotron.org.uk
Levels of processing (LOP)

Different levels of processing:
Structural – appearance
 Phonological – auditory/sound
 Semantic – meaning


Structural is the shallowest, semantic is the
deepest
www.psychlotron.org.uk
Levels of processing
Shallow processing
Structural
Weak memory trace
Deep processing
Phonological
Semantic
Strong memory trace
www.psychlotron.org.uk
Levels of processing
Incoming stimuli pass through a series of
analysing mechanisms
 Memory traces are a product of how stimuli
are analysed
 Strength of trace depends on:

Attention paid to stimulus
 Depth of processing carried out
 Connections with existing knowledge

www.psychlotron.org.uk
Levels of processing
The basic prediction of the LOP approach is
that the amount of info P’s will
recall/recognise will depend on how deeply
the experimental stimuli were processed
 Complete experiment into LOP

www.psychlotron.org.uk
Levels of processing

Elias & Perfetti (1973)

www.psychlotron.org.uk
Supporting evidence
P’s had greater recognition of words they had
thought of similes for (semantic) than word they
had thought of rhymes for (phono)
Supporting evidence

Craik & Tulving (1975)

1.
2.
3.

Tested P’s in 3 ways:
Is the word written in
capitals? BIRD (y/n)
Does the word rhyme
with ‘love’? Dove (y/n)
Complete the sentence
… ‘the man ate the …
telephone/apple.’
Highest recognition of
semantically processes
stimuli, followed by
phono, followed by
structural
70
60
50
40
structural
phonetic
semantic
30
20
10
0
%
www.psychlotron.org.uk
Levels of processing - positives




Influential model that focused researchers on processes that
they had tended to neglect
The idea that the nature of a memory trace depends on
encoding processes is well supported by empirical evidence
Theory can be applied to everyday life, e.g. exam revision
You will recall more if you use…
 Depth – make sure you understand & make connections
between the topics & ideas
 Spread – use several different techniques on the material
 Elaboration – mental effort is required to store material
effectively
 Distinctiveness – make the material your own

Many different variables involved in
determining how a stimulus is
processed:
Depth
 spread
 Elaboration
 Distinctiveness


Very difficult to isolate these variables
experimentally
www.psychlotron.org.uk
Levels of processing - negatives
Semantic processing was not always best;
it depended on how recall was measured
and how relevant the task was (Morris et al
(1977):
 Recognition – semantic best
 Rhyming recognition – phonological best
 It is difficult to measure ‘depth’
 P’s may not process the info the way you
want them to
 The theory ‘describes’ rather than ‘explains’

www.psychlotron.org.uk
Levels of processing - negatives
Explanations of Forgetting
ACTIVITY – group work/presentations
Trace decay
 Displacement
 Lack of consolidation
 Interference
 Retrieval failure
 Motivated forgetting (repression)
