Transcript 10 - smw15.org

PSY 445: Learning & Memory
Chapter 10:
Storage & Retrieval
LTM for naturalistically learned material
Bahrick, Bahrick, & Wittlinger (1975)
• Recall
• Can you list all your classmates?
• Can you name all these faces?
• Recognition
• Is this the name of a classmate?
• Is this the face of a classmate?
• Match these names and faces
Tested nearly 400 high-school graduates on their
ability to recognize and name classmates after
delays of up to 30 years.
Bahrick, Bahrick, & Wittlinger (1975)
Results
• Recall
Ability to recall a name:
• 15% for recent grads
• 7% for older participants
• Picture Recognition Test
• 50% recall after 34 years
• Name Matching Test
• 75% recall at 34 years
• 60% recall after 47 years
Interpretation
• The memory for faces of high school classmates seems quite durable
• Forgetting is rather gradual
• Recall, but not recognition, of well-learned personal material, closely
follows the forgetting curve first demonstrated by Ebbinghaus (1913)
Remembering Knowledge Learned in
School
Semb, Ellis, & Araujo (1993)
• Assessed learning 4 and 11 months after participant
took a child psychology course
• Grades fell only about 20%
• MC > Recall
Conway, Cohen, & Stanhope (1991)
• Former cognitive psych college students tested up to
12 years after completion of the course
• Largest drop occurred with the first 3-4 years; then
gradual forgetting afterwards – just like Ebbinghaus
Remembering Knowledge Learned in School
Bahrick (1984)
• Tested individuals decades after taking Spanish or
algebra
• Could tell who got the good grades and who didn’t
do so well
Memory for TV shows
Squire & Slater (1975)
• MC format of titles of the shows
• Recent shows (one to two years) = 70%
• Older shows (eight to fifteen years) = 55-60%
Memory for public events
Marslen-Wilson & Teuber (1975)
• Older individuals were better at identifying faces from
the more recent decades and forgot more of the names
from earlier decades
Long-Term retention in animals
Vaughn & Greene (1984)
• Pigeons were able to remember discriminative
stimuli for food availability up to 730 days
The Nature of Storage
Psychological Models of Semantic Memory
Basic set-up:
• Nodes represent concepts in memory
• Relations represented links among sets of nodes
• Spreading activation
Robin
Property
Wings
Collins & Quillian’s Model (1969)
Spreading activation
• Activation is the arousal level of a node
• Spreads down links
• Used to extract information from network
Structure is hierarchical
• Time to retrieve information based on number of
links
Inheritance
• Lower-level items also share properties of higher level
items
Collins & Quillian’s Model (1969)
Animal
Has skin
Breathes
Eats
Has fur
Has fins
Fish
Salmon
swim
Has gills
Dog
barks
4 legs
Is pink
Has spots
Dalmatian
Is edible
Lays eggs upstream
Skinny tail
Black &
white
Collins & Quillian Model (1969)
Support
• Collins & Quillian (1969) serial reaction time (SRT)
experiment
Criticism
• Some violations in SRT experiments
• Does not explain the typicality effect
Models of Semantic Memory
Collins & Quillian (1969)
Evidence: When people do a semantic verification task,
you see evidence of a hierarchy (response times are
correlated with the number of links)
Superset (category):
Property:
Number of links:
A canary is a canary
(S0)
A canary can sing (P0)
0
A canary is a bird (S1)
A canary can fly (P1)
1
A canary is an animal
(S2)
A canary has skin (P2)
2
Models of Semantic Memory
Problems with network models:
Violations in hierarchy
• With the hierarchy: “A horse is an animal” is faster than
“A horse is a mammal” (which violates the hierarchy)
• A chicken is more typical of animal than a robin, a robin is
more typical of bird than a chicken.
• How can a network account for this?
Answering “no”
• You know the answer to a question is “no”(a bird has four
legs) because the concepts are far apart in the network
• But, some “no” responses are really fast (a bird is a fish)
and some are really slow (a whale is a fish).
• The reason for this isn’t obvious in the model
Problems with network models
Do not explain the
typicality effect
Typicality Effect
Smith et al. (1974)
Statements about prototypical objects are verified quickly
• Is an apple a fruit?  Yes/No
• Is pomegranate a fruit?  Yes/No
Collins & Loftus (1975) Semantic Model
• Got rid of hierarchy
• Allowed links to vary in length to account for
typicality effects
See Next Slide 
Collins & Loftus (1975)
swims
fish
4 legs
Goldie
fur
dog
pet
Lucy
mutt
poodle
Neuropsychological Dissociations
A modular approach suggests that there are
hypothetically separate memory modules in the brain,
each underlying different sorts of knowledge
Caramazza & Hillis (1991)
• Two patients were impaired at verb production
• One had problem with reading; the other with writing to
dictation
• Verbs would be mispronounced; misspelled, misread, or go
unrecognized as words
• Presented homonyms
• Patients could read word when it was a noun but not as a verb
• “There was a crack in the mirror”– no problem
• “Don’t crack the nuts in here” – could not read this
Biological Substrates
Engram
• Neural representation of the brain
• Ongoing search in the field of learning and memory
Theories
• Memory-molecule theory
• Synaptic connection theory
• Long-term potential
Biological Substrates
Memory-Molecule Theory
• Much evidence that biochemical changes occur
during the brain during learning
• This theory posits that a unique memory molecule is
formed that encodes each new experience
• Early hopes that we could develop a “knowledge pill”
have faded
• Early support for this theory have faded as well
Biological Substrates
Synaptic Connection Theory
• Much evidence that the number of synaptic connections
between neurons is not fixed but is affected by life
experiences
• Long-term potential in hippocampal cells may lead to better
memory
• This theory posits that: repetitive stimulation in one cell (or a
set of cells) enhances the capacity for adjacent cells to be
triggered
• Spreading activation occurs from one cell to another
• Repeated experience allows activation of one cell by another
to occur more readily
• One stimulus cues another as in classical conditioning, pairedassociate learning, and cued recall
Synaptic Connection Theory
Kandel (2001)
• Found evidence for two storage process: STM, LTM
• Synaptic changes seen that underlie learning in a marine snail
(Aplysia)
• Long-term potentiation also seen in experiments on mice
• Transient chemical changes occur at the synapses that allow
subsequent stimuli to trigger nerve impulses
• This process lasts for hours
• Repeated intense stimulation leads to chemical changes in the
synapse, coding in the nucleus of the cell and the growth of
additional connections between cells
• These final changes are the ones that characterize
consolidation and the formation of durable LTM
Consolidation Theory
Memories start off in temporary form and over time
become consolidated into a more permanent form
• Reverberation can occur
• The initial STM’s might be retained in the brain in one
form (possibly in the persisting activity of recently
triggered nerve cells)
• These transient memories are susceptible to disruption
or interference which can block their consolidation into
permanent memory
• LTM might be encoded by structural changes that occur
at the synapses
• The consolidated memories are more resistant to loss,
change, or interference
Consolidation Theory
• This theory posits that when learning occurs it can
be demonstrated immediately after the learning
experience
• However, the memory is in an unconsolidated state
(STM processes) which will decay and be forgotten
unless consolidated into permanent memory
Testing for Consolidation
Consolidation Blocking Experiments
Electroconvulsive shock (ECS)
• Given shortly after a learning trial prevents retention
of that learning
Protein-Synthesis Inhibitors
• These drugs inhibit metabolic behavior in the brain
shortly after the learning trial
• The normal chemical activities that could possibly
have consolidated a fresh memory are blocked;
permanent memory is not possible
Retrieval from Episodic Memory
Distinctiveness
• Retrieval cues may be uniquely targeting a single memory
• Flashbulb memories
• Von Restorff effect
Testing Effects
• Retrieval is facilitated by previous retrieval
• Improvement in recall performance arising from repeated
testing sessions on the same material
• Increases with more recall tests
• Largest effects on free recall, but also appears for cued recall
and recognition
• Doesn’t fit in with forgetting paradigm
See next slide 
+ Testing Effects
Erdelyi & Kleinbard (1978)
Procedure
• Studied 60 object line drawings (pictured objects
such as DOG, CAT, or TREE)
• Only one study period to learn words
• Tested 20 times over the following week
• Repeatedly tried recalling the pictured objects
+
Has Ebbinghaus Decayed with Time?
Erdelyi & Kleinbard (1978)
Results
• Recall improved over testing days
• Essentially reversed Ebbinghaus’s forgetting curve
Interpretation
• Hypermnesia arises through visualization and reconstruction
Hypermnesia in Everyday Life
Bluck, Levine, & Laulhere (1999)
Procedure
• Asked people who had viewed the televised OJ
Simpson verdict 8 months prior to recall details
about the event
• Participants were interviewed three times in a row
Results
• The number of verifiable details remembered
increased over the interviews, revealing
hypermnesia
• Verifiable details went from 27% to 52% with no
notable increase in % of false alarms
Hypermnesia: Why the better recall?
Why do people display better memory across
successive tests?
• More time to retrieve
• Self-prompting cues vary across tests
• More chances for encoding and retrieval
processes to match
Implications for eyewitness testimony
• Maybe first recall isn’t the most valid
• Hypermnesia effects would suggest that
additional accurate information can be
ascertained through repeated recall attempts
What makes a good retrieval cue?
• Associations
• Encoding Specificity
• Contextual Learning
• State-Dependent Learning
• Mood-Dependent Recall
Associations
Effective retrieval cues include those that are
connected to the target by strong associations
Mantyla (1986)
Procedure
Acquisition Phase
• Participants studied 600 words; as each word was presented,
participants were asked to generate up to three properties
that described the word
Test Phase
• Either the self-generated one-word descriptors or someone
else’s cues were supplied to the participant when attempting
to recall a particular word
Note: Total time for both phases was 6 hours (one session)
Associations
Mantyla (1986)
Results
Immediate Recall Test:
• 90% recall with own words
• 20% recall with someone else’s words
One-Week Delayed Recall Test:
• Approximately 65% recall with own words
• 20% recall with someone else’s words
Encoding Specificity
Tulving & Thomson (1973)
• Theoretical interpretation of relationship between info at
encoding and cue at retrieval
• Specific encoding operations determine what is stored,
and what is stored determines what retrieval cues are
effective in providing access to it
• Recognition occurs when the memory test reactivates the
same meaning with which target words had been
encoded (during acquisition)
• Specific encoding operations determine what is stored,
and what is stored determines what retrieval cues are
effective in providing access to it
Matching Conditions of Encoding & Retrieval
Encoding Specificity Principle
Godden & Baddeley (1975)
Procedure
Participants randomly assigned to 4 groups
Listen/Recall
Group 1:
Group 2:
Group 3:
Group 4:
water/land
land/water
water/water
land/land
37
Matching Conditions of Encoding & Retrieval
Procedure & Results of Gooden & Baddeley (1975)
38
Matching Conditions of Encoding & Retrieval
Procedure & Results of Grant et al. (1998)
Matching Conditions of Encoding & Retrieval
Interpretation of both studies:
• The effect of context is significant
• Recall is better if the retrieval context is similar to
the encoding context
Gooden & Baddeley (1975); Grant et al. (1998)
State Dependent Learning
Memory is better when a person’s internal state
during retrieval matches their internal state during
encoding
• Information learned in the same drug-influenced
state leads to better memory
Mood Dependent Recall
• Moods Cue
• Memory is better when a person’s internal state
during retrieval matches their internal state
during encoding
• Information learned in a particular emotional
state (e.g., depressed, happy, etc.) may be more
easily recalled when in that same state of mind
Murray et al. (1999)
Mood congruence
See next slide 
Mood Dependent Recall
Murray et al. (1999)
43
Mood Dependent Recall
Procedure & Results of Eich & Metcalfe (1989)
Mood Dependent Recall
Interpretation
• The effect of mood state is significant
• Recall is better if the retrieval mood state is similar
to the encoding mood state
Emotional Arousal and Retrieval
High levels of emotion can block retrieval of memories that
otherwise could be recalled
Berkun et al. (1962)
Procedure
• Army recruits tested in apparent life-threatening situation or
non-threatening situation
Acquisition Phase
• Before boarding told to study emergency instructions
• During flight distress group’s engine stops working and they
are told to prepare for evacuation into the ocean below them
• Non-distress group had nice flight
Emotional Arousal and Retrieval
Berkun et al. (1962)
Procedure
Test Phase
• Complete form noting personal possessions
• Given written test of emergency procedures
Results
• Distressed group recalls significantly fewer words than the
non-distressed group
Interpretation
• High emotions can block retrieval of memories that would
otherwise be recalled
Feeling of Knowing (FOK)
Occurs when a person is convinced that they know
something, yet is irritated and baffled by their
inability to demonstrate this
• Similar to tip of the tongue (TOT) except TOT is more
intense as the person is sure they are close to recall –
not the case with FOK
• FOK example – I know I know their name but I am
clueless at the moment
• TOT example - I know I know their name and I’ll get it in
a second
False Retrieval
The mistaken recall of some stimulus or event that
had not actually occurred
DRM procedure
• Lists of words are presented in which all the words
are associated with a target word that was not
included in the list
• During testing, many participants produce false
recall of the target word
• Deese (1959): 30-40%
• Roediger & McDermott (1995): 50%
Source Memory
Distinguishing between actual experienced events and those
events that were imagined, thought of, heard about, or even
dreamed
• Forgetting the source of memory can lead to the false belief
that the event really occurred
Loftus & Pickrell (1995)
• Implanted childhood memories in people
• One 14-yr old was duped into thinking he had been lost in the
mall
• Family members gave many details of fabricated event
• After two days, boy was adding new specifics
Whose memory is it?
This is a memory whose origin is claimed by two
people
Sheen, Kemp, & Rubin (2001)
• Two-thirds of same sex twins claim to have this
Imagination Inflation
Confusion between memories of actual events and imagined
events
Goff & Roediger (1998)
Procedure
• Participants were asked to perform certain behaviors and to
imagine performing other behaviors
• Later participants had to recall whether they had actually
performed the action or had simply imagined it
Results
• The more frequently they had imagined the behavior, the more
likely they were to think they had actually performed the
behavior
• They referred to this as “imagination inflation”
Interpretation
• Imagining can appears a powerful cause of source amnesia
The Effect of Postevent Information
The Misinformation Effect
• After exposure to subtle misinformation, many
people tend to misremember
• Can change how a witness describes the event at a
later date
• This misleading information after a person witnesses
an event is referred to as misleading postevent
information (MPI)
• As memory fades with time, the injection of
misinformation becomes easier
Explanations for recalling MPI
1. Original memory has disappeared; it has been
overwritten by the distorting information
received later
2. The memory of the first event has simply been
forgotten so the postevent is all that is available
for recall
3. Both memories are still there and we choose MPI
because we think it is the correct response;
makes the most sense, etc.
Reconsolidation
• This theory posits that when a memory is retrieved,
it returns to the STM state
• If undisturbed it will undergo consolidation again
back into LTM
• However, it may not go undisturbed and is
susceptible to alteration at this time
• Since reactivated memory is in a labile state, it can
be blocked (just as in consolidation)
Recovered Memory
Repressed memory or false memory syndrome?
• Much debate
• Neutral stand is to refer to these as recovered memories
Brewin & Andrews (1998)
• Offered several explanations for repressed memories
1. Unconscious distress produced by the trauma
2. Repression could be conscious and deliberate
3. Might involve implicit memory systems rather than
explicit systems
4. Adequate retrieval cues are not present (state
dependent memory idea)
Retrieval vs. Reconstruction
The constructive nature of memory
• Maybe more to retrieval than just cues, etc.
• Memories are not accurate records of what
happened but construction of what might have
happened or reconstruction
Bartlett(1932)
• Participants read a tale “War of the Ghosts” and
then re-told it several times
• Looked at progressive changes in what participants
remembered about the story
Bartlett (1932)
Omissions
• Poor recall for many of the details (specific names,
or events)
• Minor events were omitted (recall for main plot and
sequence of events was not too bad)
• Shorter than the original
Normailizations
• Tendency to add and alter the stories to make them
more conventional or reasonable (top-down
processing)
Bartlett (1932)
Human memory for this type of material is NOT
reproductive
• Not simply a passive recoding of events
Rather, it is reconstructive
• Altered during BOTH storage and retrieval
• Combining elements from the original material
with existing knowledge
• We often obtain a highly abstracted version of a
conversation that is retained; we are likely to
paraphrase the conversation when recounting it
later
• Reconstruction has taken place
Credits
Some of the slides in this presentation prepared with the assistance of the
following web sites:




psychology.illinoisstate.edu/jccutti/psych368/.../9.memory12.p...
capone.mtsu.edu/wlangsto/CogNotes8.ppt
www.sussex.ac.uk/.../Seminar5-Developmental%20Prosopagn...
www.csupomona.edu/.../PSY335%20PPTs/.../BChap10.ppt