Psy 151

Cognitive processes 3 :
Memory basic models.
Forgetting

1




Everything in life is memory,
save for the thin edge of the present.
- Gazzaniga (2000)
2


without the capacity to remember and
to learn, it is difficult to imagine what life
would be likeWithout memory, we
would be servants of the moment.
Civilization itself is the distillation of human
memory.
- Blakemore
3
4

Cognitive processes 3 :
Memory basic models; Forgetting

Basic definition of Memory:
. the processes involved in retaining,
retrieving, and using information about
stimuli, images,events, ideas, and skills
after the original information is no longer
present. (Goldstein)







In the broadest sense, learning is the
acquisition of knowledge and memory is
the storage of an internal representation of
that knowledge.
(Blakemore, 1988)
5
6

Memory is studied in terms of
Remembering and Forgetting
Different aspects/ components/ kinds of
memory have been studied/ proposed.

Retrospective and Prospective memory;
Implicit and Explicit memory ;
Sensory - Short-term Long-term memory
Long-term memory - Procedural
Declarative : Episodic & Semantic

7

Retrospective and Prospective memory :

Retrospective memory is that which
pertains to PAST experience (the more
common meaning of memory)

Prospective memory is that which pertains
to the FUTURE - it could be in terms of
TIME, or EVENT.

8

Retrospective/ Prospective memory
Retrospective memory memory of the
past;
Prospective memory memory for the
future.
Implicit / Explicit memory :
Implicit memory (or memories) lie
outside conscious awareness.
Explicit memory (or memories) are those
that can be actively brought into
conscious awareness.


Memory- its sub-processes
How does the process of remembering
take place ?
The process of memory is best
understood in terms of three sub-
processes :
Encoding taking in information,
Storage,
Retrieval getting information out of
storage.
9

Components (according to the
Atkinson-Shiffrin model)
Storage : Sensory memory
Short-term memory (STM)
Working memory (WM)
Long-term memory (LTM)
10
11
Sensory, Short term, & Long-term memory:
The most common model followed in
the study of memory (the Atkinson-
Shiffrin model, dual-store, or MODAL
model):



Selective
attention
Not coded
in STM

Incoming
information
Sensory
memory
Not attended to
Forgotten
Short term
memory
Forgotten/displaced

Successfully
Coded for
LTM storage

Rehearsal

Long-
term
memory
12

Sensory memory :
Processing similar to attention.
Distinction between
iconic memory (visual) about 0.5 sec
echoic memory (acoustic) up to 2 sec.

Has a very short duration.
Processing seems to focus on sensory
features experimental findings
show special features of iconic and
echoic memory.


Iconic memory whole-vs.- partial report
experiment (Sperling).
This finding shows that items in iconic
memory fade very fast.
If another stimulus is presented
immediately after the first one, masking of
the previous stimulus occurs (some features
of the previous stimulus are erased).
This also involves fading of the previous
stimulus components.

13

Echoic memory A special feature of
echoic memory is seen in the suffix
effect:
Imagine yourself a research participant in a
memory experiment , where a list of random
digits, letters, or the like is being presented to
you. If the list is presented to you auditorily
(as opposed to visually), and if there is an
auditory cue such as a spoken word or
specific item, recall of the last few items on
the list is seriously hindered. The recall cue is
called the suffix . (from Galotti, 2000,p. 134)
14

Short-term Memory (STM)
Gross psychologists have examined
STM in terms of
capacity limited, but can be increased
[chunking]
duration- very short- longer than
sensory memory- can be increased
[ rehearsal ]
coding mainly visual and acoustic

15

Capacity - the limited capacity of STM
can be increased through Chunking :
information is stored in units (chunks),
each with a size of 7 + 2 (magic
number).

Duration Duration in STM can be
increased through rehearsal.

Coding mainly Visual or Auditory.
Some findings also show semantic
coding (e.g. Shulman, 1970)


16
Another feature of STM
The Serial position effect the primacy-
recency effect.
The first and last part of the material
presented are recalled better than the
middle part.
The primacy effect is slightly stronger
than the recency effect. But if certain
steps are taken, the primacy and recency
effects can be equally strong.
von Restorff effect ANY part of the material
that has special features will be recalled
more easily.



17
Retrieval from STM involves a search,
which may be in the form of
serial search, or
parallel search.
In general, parallel search is more
economical in terms of time and effort.
The nature of the search depends
on the nature of the material.
If information that enters STM is not needed
for long-term storage, it is held in STM for
a short duration, and then forgotten (lost),
or displaced, to make space for new
information.

18
19

Ques : Exact nature of information
processing in STM ?
Is STM only what has been described in
the Atkinson-Shiffrin model ?

Ans. Evidence of various kinds indicates
that probably processing in STM is
more complex than what was known
on the basis of earlier research.
The WORKING MEMORY model was
proposed to provide a more
comprehensive understanding of STM.





Initial proposal of WORKING MEMORY
model (Baddeley & Hitch).
Contains the following components :
Central executive .
Phonological (articulatory) loop,
dealing with auditory/acoustic
information;
Visuo-spatial scratchpad (sketchpad),
dealing with visual information;
20
21
Model of Working Memory
(Baddeley & Hitch)

Visuo-spatial
sketchpad
CENTRAL
EXECUTIVE
Sensory Store

Articulatory
Loop
Episodic
memory
Semantic
memory
Procedural
memory
22
Central executive - the controller, or
central processing unit.
Involves ATTENTION,
Decides which information will be
focused on, and how,
Decides allocation of limited mental
resources to cognitive tasks.
Articulatory /phonological) loop deals
ith speech- and sound-related aspects ;
Visuo-spatial sketch-pad (scratch-pad)-
deals with visual-spatial aspects.

More recent model:
23
24

Long- term Memory (LTM)
Capacity : Limits not known, cannot be
specified.
Duration : Again, cannot be specified, but
is much longer than in STM.
Coding : Mainly, semantic code.
Evidence also for visual code (imagery).
LTM processes as well as storage is
very complex.


Main features of LTM:
1. Information-processing and storage
seem to be different, and more complex
in LTM compared to STM.
LTM helps in understanding
how information is stored over long
periods of time;
the role of memory in language-
related processes,
the role of language in memory.

25
26

2. Information that is to be transferred to
LTM from STM is REHEARSED - two kinds
of rehearsal :
o Maintenance Rehearsal
o Elaborative Rehearsal
3. Storage is mainly in the form of the
semantic code it can be represented in
the form of images.
4. The dual coding theory(Paivio) : proposes
that there are two subsystems for
processing meaning verbal, and non-
verbal.

27

5. LTM consists of EPISODIC, SEMANTIC, and
PROCEDURAL MEMORY.
EPISODIC and SEMANTIC MEMORY are
declarative, whereas PROCEDURAL
MEMORY is not.

6. Storage in LTM is highly organized
systematically, and hierarchically.
Associations form an important basis
of hierarchical organization.


28

7. LEVELS-OF-PROCESSING approach
(Craik & Lockhart) :
Level-wise processing is found along
with hierarchical organization
Level 1: Perception
Level 2: Structural features are analyzed
Level 3: The meaning is analyzed

8. Processing (and retrieval) is influenced
by CONTEXT-DEPENDENT MEMORY and
STATE-DEPENDENT MEMORY.

29

Context-dependent memory
Much of the information is encoded in
specific contexts - later recall of information
is better when it is done in the same
context as the one in which information
was encoded.

State-dependent memory
Recall of specific information is better
when done under the same state of
consciousness as that which the information
was taken in e.g. sober versus drunken
state, normal state vs. hypnotic trance).
30

9. LTM processes indicate the role of
other sub-processes in memory-
CONSTRUCTION and RE-CONSTRUCTION
- aspects usually not found in STM.

Construction and Re-construction refer
to the processing and storing of
information, and/or specific experiences
as construed, or made meaningful, in
the light of an individuals existing
stored information (e.g. schemas and
prototypes).
31

10. That is, information is NOT mechanically
stored in LTM; rather, it is modified in the
light of existing information - often
showing the effects of associations,
motivation, and personality.
Classic study by Sir Frederic BARTLETT.
Some findings:
Change in length of a narrative.
Filling in gaps, as well as omissions.
Levelling, and Sharpening.
Primacy and recency effects some
evidence.
32

11. RETRIEVAL , or bringing back of
information from LTM - also a complex
process.
Commonly experienced difficulty in
retrieval: the TIP-OF-THE-TONGUE (TOT)
phenomenon (Brown & McNeill).
Association of different kinds - key factor
underlying retrieval.
More details about LTM can be known
by studying FORGETTING.
Ques: Does Retrieval necessarily require
activation of STM before actual responding ?
33

Putting together information regarding
Sensory memory, STM, and LTM, some
broad inferences can be drawn:
Storage and Retrieval assume Encoding.
Storage is necessary but not sufficient
for Retrieval.
Storage in LTM also involves complex
coding.
Retrieval from LTM involves a process that
is different from that in STM.


34

FORGETTING
Forgetting is the inability to recall or
remember information that has been
encoded and stored.

Ques:
1) WHY does forgetting take place ?
2) Is information that is once encoded
and stored, ever, or never forgotten ?


35

Why does Forgetting take place ?
Several explanations have been proposed:

1) Encoding failure the information was
not taken in properly, in the first place.

2) Encoding specificity - If encoding was
done in a specific context retrieval would
be difficult if there is a mismatch between
the cue for retrieval, and stored information
in terms of the context also for a
mismatch between the original and recall
state.




.



3) Interference -
Retroactive interference/inhibition - new
information interferes with the recall of
previous/ existing information.
Proactive interference/inhibition- old
information interferes with acquisition/
retention of new information.
Ques: Which of the two kinds of interference
retroactive, or proactive - is found to
be stronger in a particular situation ?
36

Experimental procedure to compare the
two kinds of interference (before-after
experimental procedure):
A randomly selected sample is randomly
assigned to four groups,
Group IA, Group IB, and
Group IIA, Group II B.

37
38

All groups are asked to memorize lists of
nonsense syllables (e.g. Consonant-Vowel-
Consonant [CVC], CCC, etc.)
Grp IA (Exptal RI) :
Learn Learn Rest Asked to
List A List B interval recall List A

Grp IB (Control RI):
Learn (Other Rest Asked to
List A activity) interval recall List A
continued.
39
Grp II A (Exptal PI)
Learns Learns Rest Asked to
List B List A interval recall List A

Grp II B (Control PI)
(Other Learns Rest Asked to
activity) List A interval recall List A
----------------------------------------------------------------
When the relevant groups are compared,
the relative effects of RI and PI can be
known.
Ques: Specifically, which groups should be
compared ?



4) Trace decay - when information is
encoded, it is stored in the form of
memory traces (neural representations).
Forgetting occurs because these memory
traces fade with time, if they are not
used.
This idea is currently applied more to STM,
and is accepted in a modified form for
LTM.

40
41
5) Motivated forgetting :
Forgetting occurs because one is
motivated to forget. Two views/ forms of
motivation
o Repression (Freudian view) : we forget
because the to-be-remembered information
is threatening : this information is repressed,
or pushed down into the unconscious
experienced as forgetting.
o The Zeigarnik effect : Information that
arouses tension is LESS LIKELY to be
forgotten than information that is neutral or
non-tension-arousing.

42

Special case:
Tip-of the- tongue (TOT) phenomenon
represents a case of forgetting because
of retrieval difficulty.
Ques: What would be the best explanation
of TOT phenomenon ?


43

(Experimental evidence of Zeigarnik effect :
The recall of finished or unfinished tasks).

Notable feature - Both views make
opposite predictions :
Repression what is anxiety-arousing is
forgotten more easily.
Zeigarnik effect what is anxiety-arousing is
forgotten less easily.
44

METHODS USED IN THE STUDY OF MEMORY

1. Serial recall
2. Paired associate recall
3. Anticipatory recall
4. Recognition
5. Re-production
6. Dual task technique
In some of these methods, especially when
recognition is involved, the SIGNAL
DETECTION approach is used.

45

Some practical suggestions for better
memory :
Actively rehearse/ repeat learned material
wherever possible, rehearsal should be
elaborative, not maintenance.
Minimize interference.
Use mnemonic devices e.g. method of
loci (spatially locate pieces of information
in different parts of the physical
environment), make acrostics, or
acronyms that are personally easy to
remember.

46

Whole versus Part learning : Depending on
the nature of the material, it should be
memorized / learned as a whole, or
learned after dividing into suitable smaller
units.
If the material involves basically a single
principle, it is better learned by the whole
method.

Mass versus Spaced learning : Time spent
in one sitting should be long enough
to assimilate the learned material, but
should not cause mental fatigue.

Recommended : Short breaks between
long sessions.

*******

47