Transcript Unit 2-AnewPsy248

Modern Classic Theory
Dog
visual
features
shallow
letters
D
O
“d”
“o”
word
deep
G
sounds
“g”
dog
Dog Memories
Saw
“Lassie”
On TV
Got
Bit
In May
4-legs
FIDO
my
dog
animal
pet
barks
Cats
meanings
Knowledge
about dogs
Modern Classic Theory
Dog
Sensory memory
visual
features
shallow
letters
D
Activation
Of
shallowest
levels
O
“d”
“o”
word
deep
G
sounds
“g”
dog
Dog Memories
Saw
“Lassie”
On TV
Got
Bit
In May
4-legs
FIDO
my
dog
animal
pet
barks
Cats
meanings
Knowledge
about dogs
Modern Classic Theory
Dog
Short-term store
visual
features
shallow
letters
D
Temporary
activation
of medium
and deep
Levels
(limited
Capacity)
O
“d”
“o”
word
deep
G
sounds
“g”
dog
Dog Memories
Saw
“Lassie”
On TV
Got
Bit
In May
4-legs
FIDO
my
dog
animal
pet
barks
Cats
meanings
Knowledge
about dogs
Modern Classic Theory
Dog
visual
features
shallow
letters
D
O
“d”
G
“o”
“g”
transfer
to LTS
(add new
Structure)
word
deep
sounds
dog
Saw on
List
Tuesday
Oct 8
Dog Memories
Saw
“Lassie”
On TV
Got
Bit
In May
4-legs
FIDO
my
dog
animal
pet
barks
Cats
meanings
Knowledge
about dogs
What is long-term store?
General
Knowledge
Memory for
events in
your life
Skills
What is long-term store?
General
Knowledge
Memory for
events in
your life
Skills
Episodic Memory
•I remember the time when I forgot the
classroom combination
•I remember the time when I fell off my tricycle
What is long-term store?
General
Knowledge
Memory for
events in
your life
Skills
Semantic Memory
I know that 8 is a number.
I know that Walt Disney created Mickey Mouse.
What is long-term store?
General
Knowledge
Memory for
events in
your life
Skills
Procedural Memory
I know how to play tennis
I know how to speak
What is long-term store?
General
Knowledge
Memory for
events in
your life
Skills
Semantic and episodic memory together are
called Declarative or Propositional Memory.
They store meaning.
Forgetting from LTS
Decay theory
Memories fade with time
Interference Theory
Memories are lost as a direct result of
other learning
New learning interferes with old learning
Same issue as in extinction
decay vs. inhibition
Minami & Dallenbach (1946
Cockroach Experiment
Shock
Trained until learn to avoid shock
Normal
Activity
Sensory
Deprivation
100%
In Box
Savings
in
relearning
Normal
Activity
1
2
3
interval (hours)
8
Memory is lost due to interference
Jenkins & Dallenbach (1924) (humans)
100%
after
Sleep
(60%)
%
correct
recall
(10%)
awake
1
2
3
(hours)
8
Studies of Interference
Similarity Principle
The more similar any two events are,
the more their memories interfere with
each other
Example
B C P T V E C
(hard)
Y B S K X O P
(easy)
PI
RI
PI
RI
… Event Event Event Event Event Event …..
your life
attempt
to
recall
PI = proactive interference
RI = retroactive interference
Similarity in paired-associate learning
The retroactive interference (RI) paradigm
Experimental Group
List 1
Control Group
List 1
DAX-quickly
ZIL-happy
YOV-trouble
same items
learn until perfect
List 2
DAX-soldier
ZIL-traffic
YOV-heavy
Test for DAX-?
List 1 ZIL-?
YOV-?
rest
DAX-?
ZIL?
YOV-?
control group is better
RI experiment shows that new learning interferes
with old learning
Does old learning interfere with new learning?
YES,
Proactive Interference (PI) Experiment
Experiment
Group
Control
Group
Learn List 1
Rest
Learn List 2
Learn List 2
Recall List 2
Recall List 2
Control group does better
Release from P.I. (Wickens)
change
category
%
correct
don’t
change
category
|
1
|
2
Trials
|
3
|
4
•Proactive interference increases from trial
1 to 3
•But if items on 4th trial are very different,
recall is better
•Shows that similarity causes interference
Which changes give the biggest release from PI?
Hi-release
Lo-release
animals - furniture
2-syllable — 3-syllable
numbers - letters
ink color
“good” words - “bad” words
A change in meaning produces biggest release
What release from PI experiments tell us
•similar things interfere in memory
•the similarity that matters in recall from
LTS is semantic
Therefore, words are stored in LTS in
terms of their meaning
Do we ever really forget?
Once in LTS, does it stay?
Loftus & Loftus (1980)
Which do you believe?
(1) Everything we learn is permanently in the
mind, although sometimes particular
details are not accessible. With
hypnosis or other techniques, inaccessible
details could be recovered.
(2) Some details that we learn may be permanently
lost from memory. Such details could
never be recovered because they are no
longer there.
Are these reasons for
believing memory
is permanent?
•Penfield’s brain stimulation experiments
•hypnosis
•spontaneous recovery of memories
•psychoanalysis
From Loftus & Loftus (1980)
Loftus’s Experiment
STOP
Subjects view a
series of slides
STOP
Phase 2
Subjects asked 20 questions
“Did you see a tree?”
“What color was the car?”
“Did another car come up behind the red car when
it went through the yield sign?”
(misleading question)
Phase 3
Forced choice recognition test
A
Yield
Sign
or
B?
Stop
Sign
Results of Stop-Yield Experiment
If you got misleading question, 80% picked the
wrong slide
If you did not get the misleading question
(control group), subjects picked the
correct slide
Loftus’s Interpretation
old memory (stop sign) is replaced by
the new information (yield sign)
OLD MEMORY IS GONE
BUT…..
This experiment does not prove the old
memory is gone, only that it was not retrieved.
Is Old Memory Really Gone?
Other experiments by Loftus
Subjects told afterward that they may have been
fooled;
90% still thought it was a yield sign
Offered $25 if they got it right
Still got it wrong
Hypnosis (Putnam, 1979)
made subjects worse
McCloskey & Zaragoza’s Criticism
You don’t forget the stop sign, you fail to retrieve
it
When you fail to retrieve it you must guess on the
test
AND subjects are biased to guess “yield” because
they remember the question
STOP
“yield
sign”
McCloskey & Zaragoza’s Experiment
Phase 1
Slides - petty theft of money
man reaches in tool box under a
hammer to steal money
Phase 2
Misled Group
Control Group
Read transcript
of event that
mentions
“screwdriver”
Transcript
mentions
“hammer”
Phase 3
Half of the subjects in control and misled
group are tested on the original Loftus procedure
Was it a hammer or a screwdriver?
The misled group chooses hammer 38%
The control group chooses hammer 72%
The other half of the subjects are tested
on a modified procedure
Was it a hammer or a wrench?
misled group “hammer”
control group “hammer”
72%
75%
Original memory of “hammer” is NOT
destroyed in misled condition
Meaning
Semantics is the study of meaning.
Problem:
What is the mental representation of the
meaning of a sentence?
Color
of grass
=
“Grass is green”
Methods
Linguistics
Artificial Intelligence
Experimental Psychology
Meaning of a sentence = a set of propositions
Proposition = “a basic idea”
a unit of meaning just large enough to be
true or false
Propositions
ROSES ARE RED
MY DOG HAS FLEAS
GRASS IS PURPLE
Not propositions
DOGS
PURPLE
A proposition isn’t a sentence;
It’s a thought
“The book is red”
“The book’s color is red”
“Das buch ist rot” (German)
express
the same
proposition
Complex sentences express more than one
proposition
“The big brown dog is hungry”
1. DOG1 is BIG
2. DOG1 is BROWN
3. DOG1 is HUNGRY
“Jane and Bill are students”
Formal Treatment
Proposition = Abstract unit of meaning that
expresses a relation among
concepts
concepts
(nouns or
noun phrases)
“John”
“dogs’
“Fido”
“justice”
Relations
(verbs or verblike words
or phrases)
“gives”
“likes”
“is next to”
Concepts
Relation
IS NEXT TO (JOHN, BILL)
LIKES (JOHN, BILL)
JOHN
likes
BILL
Semantic Memory
(General Knowledge
Basic Facts
“Generic” Memory)
contains propositions
concepts
TABLE
is
CHAIR
has
relation
FLAT
BACK
Semantic Memory
Network of Propositions
Furniture
Sitting
Flat
4 legs
Table
Back
Chair
Semantic Memory
Network of Propositions
Furniture
Sitting
Flat
4 legs
Table
Back
Chair
Cognitive Economy Question
FOUND IN (CHAIRS, BUILDINGS)
FOUND IN (TABLES, BUILDINGS)
FOUND IN (FURNITURE, BUILDINGS)
How are these propositions represented in
semantic memory?
Semantic Memory
Network of Propositions
Buildings
found in Furniture
Sitting
Flat
4 legs
Table
Back
Chair
One possibility: Non-economical Storage
Store everything separately
FOUND IN (TABLE, BUILDINGS)
FOUND IN (CHAIR, BUILDINGS)
FOUND IN (FURNITURE, BUILDINGS)
Semantic Memory
Network of Propositions
Buildings
found in Furniture
Sitting
Flat
4 legs
Table
Back
Chair
Other possibility: Economical Storage
Store only what is necessary. If you can
infer something from what is already stored,
don’t store it.
STORE:
FOUND IN (FURNITURE, BUILDINGS)
DON’T
STORE
FOUND IN (TABLE, BUILDINGS)
FOUND IN (CHAIR, BUILDINGS)
Economical Storage - only what’s necessary
Non-economical Storage - store everything
animal
has
skin
is
feathers
bird
fly
is
sing
ostrich
canary
yellow
Collins & Quillian (p. 417)
SUPPORTS
Economical Storage - only what’s necessary
Non-economical Storage - store everything
Each link adds .08 seconds
animal
has
skin
“Canaries have
skin”
RT = 1.46 sec
is
feathers
bird
“Canaries can
fly”
RT = 1.38 sec
fly
“Canaries are
yellow”
RT = 1.30 sec
is
sing
ostrich
canary
yellow
Collins & Quillian (p. 417)
Conrad’s Experiment
Should be
harder
“A shark has teeth”
“A shark has gills”
animal
have
teeth
is
fish
have
gills
is
shark
But results showed
“A shark has teeth” is faster
Conclude
Cognitive economy does not occur for
frequently associated properties
Conrad’s Experiment
Should be
harder
“A shark has teeth”
“A shark has gills”
animal
have
teeth
is
have
fish
gills
is
shark
have
teeth
But results showed
“A shark has teeth” is faster
Conclude
Cognitive economy does not occur for
frequently associated properties
Semantic Memory
Knowledge of facts
canary
is
bird
Episodic Memory
Personal events
“I remember going bowling
last October”
Are episodic and semantic memories
part of the same memory system. Or are they
separate kinds of memory:
Our intuitions suggest that they are different.
The relation between
episodic and semantic memories
Events in your life
Memory
for
the specific
event
(episodic memory)
Memory for
the facts that
you learn
from events
(semantic memory)
Semantic memories are strong
(often used)
Many episodic memories are hard to retrieve
(not often used, much more interference)
Source amnesia (Thorn, 1960)
We can forget the event, but remember
the facts
Semantic Part
Amethyst
Episodic Part
I
(while
turns
Yellows
Source amnesia in hypnosis
(Thorn, 1960)
•subjects hypnotized
•given general test of semantic memory, e.g.,
“An amethyst is a blue or purple
gem. What color does it turn when
heated?”
•subject told “yellow”
subject awakened
Asked same question.
Gets it right, but doesn’t know the source
forgot episodic (event)
remembered semantic (general knowledge)
George
was
Washington
2nd
grade
teacher
said
(Jan 10 in school)
First
President
George
was
Washington
2nd
grade
teacher
Older
Sister
said
(Jan 10 in school)
said
(May 3 at home)
First
President
George
was
Washington
2nd
grade
teacher
Older
Sister
I
said
(Jan 10 in school)
said
(May 3 at home)
read
(June 10 at home)
First
President
Semantic and episodic memory are in the
same memory system.
But we often forget episodic memories
because they have more interference and are used
less often.
Thirsty
Windows
Circular
Generation
Triumph
Dilute
Scorpion
Perfume
Mortal
Gothic
Spices
Bridal
N
O
S
R
L
1. Mor
2. Wea
3. Tri
4. Col
5. Gra
6. Sco
7. Bri
8. Tan
9. Dil
10. Sca
11. Gar
12. Spi
13. Win
14. Per
15. Tra
16. Cir
17. Mar
18. Par
19. Gen
20. Dis
21. Thi
22. Ban
23. Sal
24. Got
1. Mortal
2. Weapon
3. Triumph
4. Colorful
5. Gradually
6. Scorpion
7. Bridal
8. Tangent
9. Dilute
10. Scanner
11. Garlic
12. Spices
13. Windows
14. Perfume
15. Trailer
16. Circular
17. Marble
18. Parent
19. Generation
20. Disown
21. Thirsty
22. Banker
23. Salary
24. Gothic
S
N
S
N
N
S
S
N
S
N
N
S
S
S
N
S
N
N
S
N
S
N
N
S
Implicit Memory
Changes in performance on a task due to
a study episode even though you are not
trying to remember the items
examples
Stem completion
Fragment completion
SCO
S_O__I_N
Explicit Memory
Your ability to intentionally remember
items from a study episode
examples
Recall
Recognition
Read these sentences
 THE GIRL BROKE THE WINDOW.
Test reading a week later
 THE GIRL BROKE THE WINDOW.
Result:
Backward sentences are read faster if
they were backwards the first time
implicit or explicit?
Ask subjects to judge whether each sentence is
presented the same way as before
implicit or explicit?
Processing Dissociations
Some variables affect explicit, but not implicit
memory, or vice versa
Example
Level of processing affects explicit,
but not implicit memory
Graf & Mandler (1984)
Study words
Rate “liking”
(deep)
count “enclosures”
(shallow)
Test
Stem completion
(implicit)
Cued recall
(explicit)
Processing level
has no effect
Deep processing
leads to better
memory
Roediger & Blaxton (1987)
Modality (visual vs. auditory words)
Implicit memory is reduced if
the modality is changed from
study to test
STUDY
hear “morbid”
OR
see MORBID
TEST
fragment completion
M_R__D
Fragment completion is worse when you hear the
word
Explicit memory (e.g., recall) is not affected by
modality of the words
Other Dissociations
Weldon & Roediger (1987)
Study words & pictures
DOG
PENCIL
CAR
TEST
fragment completion
implicit
free recall
explicit
P__C__
fragment completion
is better only for
studied words
pictures are
recalled better
than words
Results are opposite for implicit and explicit
Implicit memory (stem completion)
Changes in performance on a task due to
a study episode
•not affected by level of processing
•is reduced when modality of the word changes
•does not occur when studied episode involves
pictures
Explicit memory (recall, recognition)
Intentionally remembering items from a
study episode
•strongly affected by level of processing
•not reduced when modality changes
•is better when the studied episode is pictures
instead of words
Why do implicit and explicit memory test act
differently?
The tests reflect memory for different information
MORBID
word
meaning
MORBID
letters
M
O
R
B
I
D
Why do implicit and explicit memory test act
differently?
The tests reflect memory for different information
MORBID
explicit memory
add new event to LTS
word
meaning
Event
on
in
Psych
248
Nov 6
MORBID
implicit memory
strengthen
existing
connections
letters
M
O
R
B
I
D