Chapter 7

Memory
PSYC 200
The Atkinson and Shiffrin model of memory
storage
Sensory Memory
• A bolt of lightning is actually 3 or
4 separate bolts, with each
lasting only about 1 ms; there is
a separation of about 50 ms
between bolts.
• The entire lightning strike lasts
no more than 200 ms (about 0.2
second). • Why do people
• Estimated duration: a little bit overestimate the duration
more than a half second. of a lightning strike?
The individual sees an afterimage
of the original sensory input. Sensory/iconic memory
Duration of this afterimage = 0.2
to 0.3 seconds

Duration of physical
Physical stimulus
stimulus = 0.05s
disappears after
0.05s
Capacity of Short-Term Memory
• Short-term memory is also limited in the number of items it can hold.
• On average, people can hold 7 items (or 7 chunks of information) in
short-term memory.
• The range is 7 ± 2
• George Miller published a paper in 1956 with this title: The magical
number seven, plus or minus two: Some limits on our capacity for
processing information.
• How do you design an experiment to test the capacity of short-term
memory?
Other names for sensory memory
• Iconic memory
• Afterimage
Another name for short-term memory
• Working memory
Distinguishing Long-Term Memory from
Short-Term/Working Memory
• Free recall: subjects are free to recall a list of items in any order.
• Serial recall: subjects are to recall the list of items in their original
order of presentation.
• Major difference between the two recall patterns:
In serial recall: subjects have good memory for the beginning of the
list, performance is poorer toward the end of the list.
In free recall, subjects have good memory for both the beginning and
the end of the list (this is called the serial position effect).
• Why?
Long-Term Memory
• Long-term memory is an unlimited capacity store that can hold information
over lengthy periods of time (i.e., it has infinite capacity and duration).
• One viewpoint assumes that information is never lost from LTM. If you
cannot remember something, that thing is lost in memory, not lost from it.
• Failing to find something does not mean that the thing has vanished (or
does not exist).
• This is the retrieval failure viewpoint.
• Another viewpoint assumes that information in LTM may decay, and hence,
is lost from it over time. This is the decay viewpoint.
What evidence suggests that information is never lost from memory?
Look at the difference between a recall test and a recognition test:

People could recall

forgotten info when
tested differently.

People could recall

only 10% of the
learned materials.

Retention = Proportion of material remembered

Recall vs. Recognition Test
• A recall test requires participants to reproduce information on their
own without any cues
• E.g., subjects learn 25 nonsense syllables. A recall test (also called free recall
test) asks the subjects to recall as many of the syllables as they can
remember.
• A recognition test requires participants to select previously learned
information from an array of options.
• E.g., subjects learn 25 nonsense syllables and then see a list of 100 nonsense
syllables. The test asks the subjects to identify the syllables that they have
learned earlier (also called old/new recognition test).
What evidence suggests that information is
never lost from memory?
• Look at the difference between first- and second-time learning.
• Relearning of the forgotten information is faster than first-time
learning. This is not possible if the information has been lost from
memory.
• A relearning test asks subjects to memorize information a second
time to determine how much time or effort is saved by having learned
it before.
• Retention is measured as a saving score:
• E.g., It takes you 20 minutes to learn a list of words the first time. It takes you
only 5 minutes to relearn it a week later. Saving = 15 minutes
• Saving score = 15/20 = 75% (you have retained 75% of the information)
 Elaboration
Effective
Encoding
Chunking
 Why do some people have better memory?

Deep Use of visual

Organization
processing imagery

Elaboration
What make people better readers/problem
solvers?

Same answers as above

Levels of Processing
• When you are presented with a word, there are several aspects of the
word that you can attend to. For example:
• The word’s physical structure (e.g., morpheme)
• Its sound (e.g., phoneme)
• Its meaning (e.g., semantics and the deep structure of language)
• Levels-of-processing (also called depth of processing) theory proposes
that deeper levels of processing result in longer-lasting memory
codes.
• The assumption is that levels of processing increase from physical
structure to phonemic structure, to semantic structure.
Results: from
Craik &
Tulving
(1975)
 Chunking
A study by Chase & Simon (1973)

Participants: Three chess players at different chess playing levels

(master, intermediate, and beginner).
All players were given 5 seconds to study the positions of chess
pieces taken from 1) actual games; 2) random arrangements.
The players then reproduced the chess positions from memory.
Results
Summary of Results
When the to-be-remembered chess
positions were from actual games:
• The chess Master placed 16 (out of 24)
pieces correctly on his first trial.
• The beginner player placed only 4 correctly.
• The master only needed 4 trials to
accomplish 100% accurate criterion.
• The beginner still made errors on his 7th
trial.
Summary of Results
When the to-be-remembered chess positions were
randomly arranged:
• Chess master performed as poorly as the beginner.
• They both recalled only 3 positions correctly.
Conclusions:
• Chess master’s advantage was due to his ability to
group the chess pieces into meaningful chunks.
• Chess master had stored in LTM many meaningful
patterns from real games.
• He saw the 24 pieces as 4 to 6 chunks.
Take-Home Message
• All experiments in this chapter repetitively demonstrate that good
memory involves some sorts of organization.
• Things that are randomly put together are hard to remember.
• To remember random patterns, you want to make up some meanings
for them.
• Organization = chunking
• People actively organize information
• People are too ready to organize information that they sometimes see
subjective meanings in a random pattern.
Long-Term Memory Requires Consolidation

Memory consolidation: The neural process by which encoded information is stored in

memory
Brain plasticity revisited: Memory consolidation is a process in which synaptic
connection is strengthened.
When a new memory is formed, the brain is changed. How?
The postsynaptic neurons are more easily activated by presynaptic neurons.
Consolidation

Remains unchanged

Traditional views of memory: Encoded information remains unchanged

in storage.
 Reconsolidation

Reconsolidation

Memories can change each time they are retrieved. Each memory is of the
previous retrieval, not the original experience.
Reconsolidation: The Constructive Nature of Memory

• Bartlett’s “War of the Ghosts” Experiment

• The subject was given a story, or an argumentative prose passage, or
a simple drawing to read/study.
• The subject attempted a first reproduction usually after an interval of
15 minutes.
• Thereafter, the subject gave further reproductions at intervals of
increasing length.
• The most well-known story that Bartlett had used is called “War of
the Ghosts.”
Reconstructive Memory: Bartlett’s “War of
the Ghosts” Experiment
• The reproduced story is considerably shortened, mainly by omissions.
• Terminology becomes more modern, more “journalistic.”
• Subsequent reproductions become more and more coherent.
• Specific terminology was replaced by more general expression (e.g.,
canoe becomes boat; hunting seals becomes fishing).
• Proper names (e.g., Egulac and Kalama) disappeared.
• Direct speech was dropped.
Reconstructive Memory: Bartlett’s “War of
the Ghosts” Experiment
• Smaller details were forgotten, e.g., the weather; other canoes; heard
the noise of paddles; face became contorted.
• The excuse of having family ties is better remembered than the
excuse of having no arrows.
• Less salient expressions were transformed to similar representations:
- river became lake, beach, island, etc.
- the log became a rock, a tree or a bush.
- “something black came out of his mouth” became “a clot of blood
came from his mouth,” or he foamed at the mouth.
Reconstructive Memory: Bartlett’s “War of
the Ghosts” Experiment
• In repeated reproductions, subjects tended to use the same
expression to paraphrase the same part of the original description.
• Although some information from the original story was omitted in
reproductions, some new information was added, and these extra
parts do not only remain but become more elaborated in subsequent
reproductions.
• Readers keep good memory of what they actively create, not what
they read in the original story.
• There is closer resemblance between two short-distant reproductions
than between the first reproduction and the original story.
Reconstructive Memory: Bartlett’s “War of
the Ghosts” Experiment
• In some cases, readers can even produce verbatim recall of their own
sentences.
• This phenomenon is known as hypermnesia.
• But what has been forgotten continued to be forgotten.
• Readers demonstrate a strong tendency to rationalize mystical
occurrences.
• The same direction of rationalization is pursued in subsequent recalls.
• Each retelling is a re-encoding. In terms of neural activity, each
retelling is a reconsolidation.
Reconsolidation: Testing Effect
• Testing does more than assessing the outcomes of learning.
• Testing improves learning.
• Which strategy produces the best learning outcome?
• Highlighting
• Rereading
• Self-testing
• Answer: Self testing
• Each retrieval is a reconsolidation. During the reconsolidation
process, misconception is corrected/modified; correct concepts are
strengthened.
A closer look at long-
term memory
Semantic Network
Imagine that you are
now sitting in this
office, waiting to
participate in an
experiment that will
take place in another
room.
Surprise Test

Write down what you saw in the office.

Now check what you have on your list
• Desks/Tables
• Chairs
• Boxes
• Books
• Filling cabinets
• Bottle
• Basket
• Umbrella
Evaluating your answers
• If you have desks (or tables) and chairs, these are “hits.”
• If you have books or filling cabinets, these are “false alarms.”
• If you don’t have bottle, basket, or umbrella, these are “misses.”
• The actual experiment (Brewer & Treyens, 1981)
• 30 subjects were tested
• Most subjects recalled desks and chairs.
• Few subjects recalled the wine bottle or the picnic basket.
• Nine subjects falsely recalled that the office had books.
Schemas
• A schema is an organized cluster of knowledge about a particular
object or event abstracted from previous experience with the object
or event.
• A typical office usually has desks, chairs, books, pens, paper,
computers (today’s standard). This is the office schema.
• People are more likely to remember things that are consistent with
their schemas than things that are not.
• This explains why most people remembered the desks and chairs,
some falsely remembered seeing books, and few remembered the
wine bottle and the picnic basket (because these items are not
consistent with the office schema).
Alternative Explanations
• Do you agree with the explanations based on schema?
• If not, how do you explain the results?
Question

Can a person have

If yes, then the two
only semantic
types of memories
memories but no
are not the same.
episodic memories?
Retrograde Amnesia
• Retro: backward
• Grade: gradient
• Amnesia: the loss of memory or memory abilities caused by brain
damage or disease.
• Retrograde amnesia: loss of memory for events before brain injury.
The loss displays a gradient across time—memories that are more
distant in time from the brain damage are less impaired by the
amnesia.
Anterograde Amnesia
• Anterograde amnesia: disruption of memory for events occurring after
brain injury.
• Anterograde amnesia often seems more extensive because it disrupts
learning from the time of the brain damage.
• A read-only memory
Real Clinical Cases
Patient K. C.
• Suffered from a serious brain injury in the frontal regions in a
motorcycle accident.
• Shows a complete loss of episodic memory—keeps no
autobiographical knowledge.
• Cannot remember any personally experienced events from both
before or after his accident—has profound retrograde and
anterograde amnesia.

50
Patient K. C.
• K. C.’s semantic memory is intact.
• He cannot remember episodic events but can remember facts about
some events (e.g., he could not remember that he had a brother but
said that his brother’s funeral was sad because he knew that funerals
are sad events).
• His episodic memory is damaged, but his semantic memory is not
affected by the damaged parts of the brain.
Patient H. M.
• Suffered from severe epileptic seizures.
• Had a part of his temporal lobes and his hippocampus in both the left
and right hemispheres removed surgically.
• Outcome: pervasive anterograde amnesia.
• Unable to learn or recall anything new.
• But memory of events before the surgery remains intact (no evidence
for retrograde amnesia).
• He could talk about his childhood and family.
• He could explain the rules of baseball.
Patient H. M.
• His I.Q. was not affected by the surgery.
• His language comprehension is normal, his vocabulary is above normal.
• However, any task that requires him to retain information across a delay
shows severe impairment, especially the delay is filled with an interfering
task.
• These impairments apply equally to nonverbal and verbal materials.
• Surgery had interfered with the process of storing new memory but had
not touched previously stored memories.
• What will happen when someone can only read, but not write anything on
LTM?
Patient H. M.
• The researchers who worked with him for 40 years had to introduce
themselves every time they met.
• He could hold a normal conversation but forgot the conversation in a
minute or less.
• He kept talking about the same subject/event without knowing that
he had already talked about it.
• H. M. became the research subject in hundreds of tasks, and he didn’t
get bored!
• In his words, “every day is alone in itself.”
H. M.’s Implicit Memory

H. M. Learned a mirror-drawing
task: he had to trace between
the lines of a double-star
pattern while looking at the
pattern and pencil only in a
mirror.

55
H. M.’s Implicit Memory

H. M.’s learning curve across three days 56

H. M.’s Implicit Memory
• H. M.’s performance on the mirror-drawing task showed a completely
normal learning curve, with very few errors on the third day of
practice.
• On day 2 and 3, he did not remember ever having done the mirror-
drawing task before.
• H. M.’s motor memory is normal even though he had no conscious (or
episodic) memory of having learned the drawing task that involved
motor learning.
• It suggests that motor learning is a type of implicit memory that by-
passes consciousness.
A closer look at implicit memory
Two Types of Priming

Perceptual priming (also Conceptual priming (also

called repetition priming) called semantic priming)
• Exposure to a stimulus • Exposure to a stimulus
will facilitate a future will facilitate a future
response (more quickly response to a new
or more accurately) to stimulus closely related
the same stimulus. to it.
What is this?
Testing Implicit Memory
Perceptual Priming (or Repetition priming)
• Definition: A previous encounter with information facilitates later
performance on the same information (or the fragments of the
original information), even unconsciously.
Example:
• The “elephant experiment”
• The word-stem completion task (see the next slide)
Perceptual Priming
Word-fragment Completion Task
Word Fragment Possible Words
SHA_E SHAPE SHADE SHAME SHARE
_EAK PEAK LEAK BEAK WEAK
_ ACK BACK LACK JACK RACK PACK
_ _ SET UPSET RESET ONSET

Participants see fragments of a word and try to fill in the missing letters to make the string of letters
a word that they first think of.
How this technique works?
People are more likely to think of “SHAPE” than “SHADE SHAME SHARE” when they see SHA_E if
they have seen the word SHAPE at an earlier time.
Note: there should be a control group or a norm to provide baseline measure of the likelihood that
people will first think of the target word without previous exposure to that word.
Perceptual
Priming
Testing Implicit Memory
Conceptual Priming (or Semantic priming)
• Definition: A previous encounter with information facilitates later
performance on semantically related information, even unconsciously.
Examples:
• You see the picture of a rabbit or hear the word “rabbit” first. Later you
hear the sound “hair/hare.” When asked to write down the word that you
heard, you are more likely to write down “hare” because “rabbit” and
“hare” are semantically related words.
• You fully understand what “negative reinforcement” means. But you
choose the word “decreases” to fill in the blank of this statement: Negative
reinforcement ____ the rate of responding.
Explicit Versus Implicit Memory
• Warrington & Weiskrantz (1970) asked amnesic patients to read a list
of words and then tested their free recall and recognition of the
words. The patients did worse than healthy individuals.
• However, amnesic patients’ performance on the word fragment
completion task was the same as healthy individuals (both groups of
participants were equally likely to think of a previously read word
when presented with its fragments).
• Amnesic patients did not have explicit memory of the learned words
but had implicit memory of them.
Testing Implicit Memory
• How do you know I know something that I don’t know I know?
• Answer from cognitive psychologists: Your behavior
• You are faster to respond to a relevant cue or
• You are better than chance to make a correct guess when presented with a
relevant cue
• Answer(s) from Sigmund Freud and other psychoanalysts?
Why We Forget: Interference
• People forget information because of competition from other
material.
• E.g., You met Joan for the first time in a party. Later in the party, you met
Nancy, Sarah, Pam, Katie, Jenny and many other women. You forgot the name
of the first woman (or might be others too) you met.
• Competition gets stronger when the to-be-remembered target bears
a close resemblance to the interfering item.
• Empirical studies demonstrating the theory of interference:
• Manipulate amount of interference by varying the degree of similarity
between the target and the interfering material.
• Prediction: Decreasing the similarity should reduce interference and cause
less forgetting.
 Proactive Interference

Info encoded in time 1 Info encoded in time 2 Attempt to recall info encoded in
time 2 but recall info encoded in
time 1 instead.

Retroactive Interference

Info encoded in time 1 Info encoded in time 2 Attempt to recall info encoded in
time 1 but recall info encoded in
time 2 instead.
Proactive Interference
• Proactive interference occurs when previously learned information
interferes with the retention of new information.
Retroactive Interference
• Retroactive interference occurs when new information impairs the
retention of previously learned information.
Serial Position Effect Revisited

Low recall due to both proactive

& retroactive interference
How We Remember?
• Using cues to aid retrieval
Example:
• Recognition Test
• The to-be-remembered items themselves are the retrieval cues
• A case of recovered memory
• DJ reported that following a dinner with a neighbor whom she had not seen
in many years, she suddenly recalled that this individual molested her
numerous times between the ages of 5 and 7.
• This case is very similar to a recognition test—when you see the target again,
you recall that you saw it earlier.
How We Remember?
• Context-dependent memory
• Also called encoding specificity
• A study by Godden & Baddeley (1975)
• Divers learned a list of words under water or on land and later
recalled the words either under water or on land. The encoding and
retrieval environments either matched or mismatched.
• The results (on next slide) show very typical encoding specificity
effect: Performances in the two congruent conditions were better
than performances in the two incongruent conditions.
How We Remember
• Reinstating the context of an event
• Everyday example:
• You visit a place in which you grew up (or a place where you used to live) and
recall many events that you have never thought of in years.
• A case of recovered memory
• JR reported that following viewing a movie involving sexual abuse, he
suddenly remembered incidents of being molested by a priest on camping
trips when he was an adolescent.
Reconstructing Memories
• Based on what we have learned about how we forget and how we
remember, many questions we may ask:
• To what extent can a lost memory be reconstructed (by the aid of a
retrieval cue or by reinstating the context of an event)?
• How accurate is a reconstructed memory?
• How reliable are recovered memories? How can we find out?
• Take a look at the case of George Franklin (very interesting)
• The Reality of Repressed Memories (washington.edu)
• The Myth of Repressed Memory (washington.edu)
• George Franklin - National Registry of Exonerations (umich.edu)
In 1990, a landmark case went to trial in Redwood City, California. The defendant,
George Franklin, Sr., 51 years old, stood trial for a murder that had occurred more than
20 years earlier. The victim, 8-year-old (Susan Kay Nason, was murdered on September
22, 1969). Franklin's daughter, Eileen, only 8 years old herself at the time of the
murder, provided the major evidence against her father. What was unusual about the
case is that Eileen's memory of witnessing the murder had been repressed for more
than 20 years.

Eileen's memory did not come back all at once. She claimed that her first flashback
came one afternoon in January 1989 when she was playing with her two-year-old son,
Aaron, and her five-year-old daughter, Jessica. At one moment, Jessica looked up and
asked her mother a question like "Isn't that right, Mommy?" A memory of Susan Nason
suddenly came back. Eileen recalled the look of betrayal in Susie's eyes just before the
murder. Later, more fragments would return, until Eileen had a rich and detailed
memory.
She remembered her father sexually assaulting Susie in the back of a van. She
remembered that Susie was struggling as she said, "No don't" and "Stop." She
remembered her father saying, "Now Susie," and she even mimicked his precise
intonation. Next, her memory took the three of them outside the van, where she
saw her father with his hands raised above his head with a rock in them. She
remembered screaming. She remembered walking back to where Susie lay, covered
with blood, the silver ring on her finger smashed.
Eileen's memory report was believed by her therapist, by several members of her
family, and by the San Mateo County district attorney's office, which chose to
prosecute her father. It was also believed by the jury, which convicted George
Franklin, Sr., of murder. The jury began its deliberations on November 29, 1990; and
returned a verdict the next day. Impressed by Eileen's detailed and confident
memory, they found her father guilty of murder in the first degree.
BURIED Series | Official Teaser - 'Repressed Memory' (HD) Showtime - YouTube

https://youtu.be/XRW-m_8nZzw
Misinformation
Effect
• Misinformation effect
occurs when people’s
recall of an event is
altered by misleading
post-event information.
• Misinformation effect is
demonstrated in a series
of studies by Elizabeth
Loftus and colleagues.
A Study by Loftus & Palmer (1974)
• Loftus & Palmer showed participants a videotape of an automobile
accident.
• After viewing the video:
• 50 subjects were asked: “About how fast were the cars going when they hit
each other?”
• 50 subjects were asked: “About how fast were the cars going when they
smashed into each other?”
• 50 subjects were not asked anything about the speed of the cars
• One week later…
• Participants were given a memory test. One question asked: “Did you see
any broken glass?”
 A study by Loftus & Palmer (1974)
• Results (screenshot of actual paper)

Subjects who had earlier been asked

about the cars smashing into each other
estimated a higher speed of the cars and
were more likely to “recall” broken glass.
Explaining Misinformation Effect
Summary
• All memories are reconstructions of the past.
• Biased or incomplete encoding leads to biased recall.
• Memory may change/lose overtime.
• Due to decay/interference
• Misleading cues in the retrieval process may also lead to biased recall.
What to read in Chapter 7?
• 7.1-7.13
• 7.17
• 7.19-7.21