FASTEXPERT
Breeuna09
ch6_7_ltm_amc_sp17.pptLong-Term Memory
Cog Labs :
Remember-Know &
Levels of Processing
Long-Term Memory
This is a huge topic - LTM is the
cornerstone of all behavior.
*
LTM….what is it?
- LTM – system that’s responsible for storing info for long periods of time
An archive of past events & knowledge
Needed for complex cognition:
Understanding language
Solving problems
Making decisions
Interacts w/ WM to create our ongoing experience
Example of WM, LTM interaction
- Hear a sentence:
- Jacob rolled the red ball down the hallway.
WM holds wording in mind
Simultaneously accesses meaning of words from LTM, which help understand meaning
LTM also has add’l info regarding those topics, people sentence refers to
Adds to understanding
Let’s begin with a
simple experiment
Please number from 1 to 15:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Leave enough
room for one
short word
after each #
Serial Position Curve
- Primacy effect – more likely to remember words at beginning of list
Subj have time to rehearse words & transfer to LTM
Less rehearsal for later words
- Recency effect – better memory for stimuli at end of list
Most recent words still in STM
Short- and long-term memory
The serial-position curve shows the
STM-LTM distinction
% correct recall
position of word in list
The primacy portion
reflects LTM
The recency portion
reflects STM
Chart1
| 40 |
| 35 |
| 30 |
| 20 |
| 14 |
| 10 |
| 9 |
| 6 |
| 4 |
| 4 |
| 2 |
| 2 |
| 8 |
| 15 |
| 25 |
| 45 |
| 55 |
| 60 |
Sheet1
| 40 |
| 35 |
| 30 |
| 20 |
| 14 |
| 10 |
| 9 |
| 6 |
| 4 |
| 4 |
| 2 |
| 2 |
| 8 |
| 15 |
| 25 |
| 45 |
| 55 |
| 60 |
Sheet2
Sheet3
- Each word passes from SM to STM
STM has limited capacity, so...
- Rehearsal transfers info from STM to LTM
- First few words make it to LTM
- Last few words still in STM at recall
Simple theory of the serial-position curve:
Long-term
memory
Working or
Short-term
Memory
Sensory
Input
Sensory
Memory
Attention
Encoding
Retrieval
Maintenance Rehearsal
Coding in ST & LTM
- Coding – form in which stimuli are presented
Visual, auditory, semantic
- Type of coding depends on task
Auditory coding is main type of coding in STM
Phone # - by repeating (auditory coding)
Not likely by visual image, meaning
Semantic coding is main type for LTM
Read a book last week & remember it
Not likely to remember what words looked like, but meaning
THE BRAIN
Working Memory, Long-term Memory, and…..
STM & LTM are separate in the B,
but some overlap. Evidence from
neuropsychology….
Milner reported the case of H.M., who
had brain surgery to relieve epilepsy...
Epilepsy improved,
but H.M.'s memory
was messed up
by the surgery.
His memory for life before surgery (LTM)
was normal… and his STM was also normal.
BUT, he lost the ability to transfer
info from STM to LTM! (hippocampus)
Lost ability to recognize people
he saw every day...
Became unable to recognize
himself as he aged.
However, H.M.
could learn.
For example,
he developed
skill in “mirror
drawing,”
without having
any memory
of learning
how…
Squire (1979) studied a similar patient
(N.A.) who had a strange accident...
The fencing foil went up his nose, and
into his forebrain!
N.A. is just like H.M.,
only a little worse...
Another example….
- Clive Wearing
Contracted viral encephalitis, destroyed parts of medial temporal lobe – including hippocampus, amygdala
Inability to form new memories, like H.M.
https://www.youtube.com/watch?v=WmzU47i2xgw
This syndrome was
the basis for the
movie
Memento
K.F.
- Damage to parietal lobe in motorcycle accident
STM was greatly impaired, but LTM was okay
Reduced digit span (2)
Could form & retain new memories
- These 3 pts (H.M., Wearing, & K.F.) show separation of STM and LTM
- Also overlap, esp for novel objects
hippocampus
LONG-TERM MEMORY
Types of
Episodic & Semantic Memory
- Episodic memory – memory for personal experiences
Main characteristic is mental time travel – “reliving” (Endel Tulving)
- Semantic memory – memory for facts, vocab, concepts
NOT tied to a specific event
Proof from neuropsychology
- K.C.
Damage to hippocampus & surrounding structures
Lost EM
E.g., knows brother died 2 yrs ago, but doesn’t recall anything personal – how he heard, felt at funeral
SM is intact
Knows where forks are in the kitchen, difference bw a strike & spare in bowling
Cont’d
- Italian woman contracted encephalitis
Difficulty recognizing familiar people
Couldn’t remember meanings of words, where things were in store
Couldn’t recall facts – who Beethoven was, Italy was involved in WW2
- Cases show 2 types of info (EM & SM) involve different B mechanisms
In real life….
- EM & SM are often intertwined
Semantic knowledge guides our experience
E.g., understanding of baseball (SM) influences EM that follows from that experience
If you know baseball, you know where to look on the field when certain things happen
If not, you miss a lot, don’t understanding what you’re seeing
Westmacott & Moskovitch (2003)
- Knowledge re public figures can have E & S
Know Oprah had a show (S)
Memory of a specific show (E)
FORGETTING
LTM (Semantic & Episodic) and….
Forgetting
- Not an ‘all-or-none’ process – degrees
More recent LTMs more detailed
Detail often fades w/ time, other experiences
Don’t remember this date 1 yr ago, but surrounding
Remember-know paradigm
- Difference for E & S memory measured via:
Familiarity (know) – seems familiar, but can’t remember details
Associated w/ SM
Recollection (remember) – remember specific details
Associated w/ EM
Petrican et al (2010)
- Examined how people’s memory for public events change over time
Presented descriptions & subj responded: remember, know, don’t know
- Results:
Complete forgetting increased over time
Remember responses (EM) decreased much more than know responses (SM)
- Semanticization of remote memories – loss of episodic detail for long ago events
FUTURE
Imagining the
Connection bw ability to remember past & ability to create future scenarios
- K.C. – lost EM
Unable to use imagination to describe personal events that might happen
- D.B.
Inability to imagine future events only applied to personal events
Could still say what might happen in politics, current events…
Addis & Schacter (2007, 2009)
- fMRI of subj remembering past or imagining future
ALL B regions overlapped!!
- Proposed Constructive episodic simulation hypothesis:
EMs are extracted & recombined to construct simulations of future events
However,…
- Some semantic dementia pts (intact EM), also have problems describing episodic details that might happen in the future
Suggests we need BOTH EM & SM to think about our personal futures
IMPLICIT MEMORY
Procedural Memory, Priming, & Conditioning
Implicit vs Explicit Memory
- Explicit memory – memories we’re aware (conscious) of
Episodic memory, Semantic memory
- Implicit memory – memory we aren’t aware of
Procedural memory, priming, conditioning
Although explicit measures are important,
we also learn a great deal from implicit
(indirect) measures.
- Implicit memories are not part of conscious awareness
- However, they enhance performance in various tasks
- Mere exposure to information can aid later perception
- May reflect a more primitive type of memory
Someone is “amnesiac” if they have
a profound memory deficit, right? Well……
- Clive can’t form new LTMs, but can still play piano
- Amnesiac pts can learn new skills even thought they don’t remember practicing
H.M. mirror drawing – trace outline while looking at its image in mirror – always thought doing it for the 1st time
Free recall
Recognition
Fragment Identification- p___um_
Stem Completion – per____
Compared amnesic patients and
age-matched controls.
All Ss completed 4 memory tests for word lists:
Warrington & Weiskrantz (1968)
Accidentally began the study of implicit vs. explicit memory.
*
W & W
results
What
The
heck
This is a very reliable finding, even
with more severe amnesiacs
Similar results also emerge in temporary amnesia,
caused by severe intoxication or electro-shock therapy!
Control subjects were told nothing about memory.
Uninformed simulated amnesics were told to pretend
they were trying to fake amnesia, but were not told that
real amnesics have normal implicit memory.
Informed simulated amnesics were told to pretend they
were trying to fake amnesia, but they were also taught
about the critical finding.
People arrested for crimes sometimes claim to have amnesia, reducing guilt by reason of incompetence.
Horton et al. compared 3 groups: subj given word
lists to remember
Horton et al. (1992) results
These people
badly over-act
Performance on implict measures
Fragment completion task
Procedural memory
- aka skill memory – memory for doing things that usually involve learning skills
E.g., tying shoes, typing, riding bike, somersault, play piano
When try to become conscious of HOW we’re doing it, we don’t do it so well!
- Associated w/
Basal ganglia – habits
Cerebellum – motor movement
Limbic system- learning, memory
This is a completely different system than
for other memory systems!! (non-hippocampus)
Priming
- Priming –when presentation of 1 stimulus changes the way a person responds to another stimulus
E..g, seeing “bird” causes you to respond more quickly to later presentation of “bird” vs. a word you didn’t see
Can occur even though you may not remember priming stimuli!
Graf et al (1985)
- Amnesiac pts w/ Korsakoff’s syndrome (can’t form new LTMs)
- Pts w/o amnesia being treated for alcoholism
- Pts w/o amnesia & no history of alcoholism
Read 10 words & rated how much they liked them
Explicit memory test – recall words
Implicit test – word completion
PER_____
Graf Results
- Amnesiac pts recalled fewer words than both control groups (explicit)
- But performed just as well on IMPLICIT measure!
Increase in performance is result of priming
Priming in everyday life
- Perfect & Askew (1994) –
- subj scanned articles in mag
Later asked to rate ads on various dimensions
How appealing, eye-catching, distinctive, memorable
- Subj gave higher ratings to ones that had been in the mag
Implicit bc subj remembered 2.8/25 ads
- Propaganda effect – more likely to rate
things as true bc you’ve been exposed to them
Classical Conditioning
- Classical conditioning
pairs a neutral stimulus that initially doesn’t result in a response &
A conditioning stimulus that does result in a response
E.g, presenting a tone followed by a puff of air to the eye, causes person to blink
- Continued association even if person has forgotten about the pairing!
CC in everyday life
- Stimuli often paired with emotions
Feel +/- about someone even though don’t remember how you know them (implicit)
- Non-implicit example
Drive by the place you got pulled over =
re-experience emotions
Isn’t implicit bc you remember what’s causing the response
Encoding
Retrieval
&
Consolidation
Chapter 7
Long-term memory
ENCODING
Getting info into LTM
Encoding
- Encoding – process used to get info into LTM
- Many ways to encode:
Maintenance rehearsal – repeating continuously. No consideration of meaning, connection
Results in poor memory
Elaborative rehearsal – relate info to something meaningful
E.g., phone #: 1st 3 same as yours, last 4 is your b-day year
Results in better memory
Craik & Lockhart (1972) –
Levels of Processing
- How you encode influences ability to retrieve
- Memory depends on the depth of processing
Shallow processing – little attention to meaning
E.g., maintenance rehearsal
E.g, paying attention to only physical features
Deep processing – close attention on meaning & relating it to something else
- Better memory for deeper processing
Craik & Tulving (1975)
- Tested memory w/ different LOP
Q about physical features of words (shallow)
Q about whether word rhymes w/ a diff word (deeper)
Q about if a word fits in a sentence (deepest)
- Memory test for words
Deeper processing assoc w/ better memory
- Depth of processing hard to define – waned, but retrieval affected by encoding accepted
The Craik & Lockhart (1972) study
Subjects asked questions to achieve different LOP.
Is the word printed in capital letters?
Does the word rhyme with train?
Does the word fit into the sentence “ He saw a ____ on the street”?
Example 1: boat: answer, no
Example 2: pain: answer, yes
Example 3: car: answer, yes
Some Real Data
LOP can be manipulated in many ways,
such as simple changes in orienting tasks…
Figure 6.9 (p. 196)
(a) Sequence of events in Craik and Tulving’s (1975) experiment. (b) Results of this experiment. Deeper processing (fill in the blanks) is associated with longer reaction times to answer the question about a word, and better memory for the word.
Surface processing
moderate processing
deep processing
Schematic example of LOP data…
Good encoding through….
- Visual imagery
- Linking words to yourself
Self-reference effect – memory is better if you relate a word to yourself
- Generating info
Generation effect – generating material yourself vs. passively receiving enhances retention
Good encoding through….
- Organization
- Relating words to survival value
Naine (2010)
Memory is shaped to increase survival value; so linking words to survival increases likelihood they’ll be recalled
May not be survival actually, but just relating words to meaningful
- Retrieval practice
Testing effect – enhanced perf due to retrieval practice
E.g., making up & answering practice Qs
RETRIEVAL
Getting info out of LTM
Retrieval
- To use encoded info, you have to retrieve it
- Important bc many memory
errors are retrieval errors
E.g., blanking on exam
E.g., can’t remember
someone’s name
Retrieval cues
- Retrieval cue – words or other stimuli that help us remember info stored in our memory
Location – return to place you originally encoded
Returning stimulates memories assoc w/ that place
Auditory
E.g., hearing a song
Smell
MATCHING CONDITIONS
of encoding & retrieval
Matching retrieval & encoding
- Retrieval can be increased by matching the conditions @ retrieval to the conditions at encoding
- 3 ways to do this…..
Encoding specificity
- Encoding specificity – we encode info along w/ its context
So, match the context (i.e., physical situation)
E.g., to recall childhood memories, go to house where you grew up
E.g., you’re in the bedroom & decide to get something from the kitchen
E.g., someone takes your seat on test day
Experimental evidence:
Godden & Baddeley (1975) had scuba divers
divers learn word lists on land or underwater...
Later tested in same or different contexts...
Godden & Baddeley’s results:
Smith (1985): Mozart or Jazz at study & test
Music at Study
State-dependent learning
- S-D learning – learning is assoc w/ a particular internal state
E.g., mood, state of awareness
Memory will be better when a person’s internal state during retrieval matches states during encoding
- Eich & Metcalfe (1989) – showed this with happy or sad mood
Matching the cognitive task: transfer appropriate processing
- Transfer-appropriate processing – better perf when type of processing matches at encoding & retrieval
E.g., rhyming, focus of meaning
dissociations between explicit and implicit memory
Time
Explicit memories decay fairly steadily over time
Implicit memories remain quite stable over long periods
Tulving, Schacter & Stark (1982)
Compared recognition
memory and fragment
completion for words
(1 hour versus 1 week)
Explicit measure
Implicit measure
Even classic findings show
dissociations between explicit and implicit memory
Levels of Processing
Explicit memories are strongly affected by “depth” of original processing
Implicit memories are relatively unaffected by LOP
Why do explicit & implicit memory differ?
Separate
brain
systems?
Different
mental
processes?
GOOD LUCK
ON YOUR EXAM!
0
10
20
30
40
50
60
70
123456789101112131415161718
5
7
9
11
13
15
DryWet
Recall environment
Mean words recalled
Dry
Wet
0
5
10
15
20
25
MozartJazz
Music at Test
Recall
Mozart
Jazz
