Transcript Lecture_16 - University of Alberta

Lecture 16 – Psyco 350, A1
Winter, 2011
N. R. Brown
Psyco 350 Lec #18– Slide 1
Outline
Autobiographical Memory
• Methods of studying autobiographical memory
• Retention Factors
• Organization
– hierarchical
– event-to-event & event cueing
• Retrieval Processes
Psyco 350 Lec #18– Slide 2
Autobiographical Memory: Methods
Cue-word Method
1. cue word  event memoryi
2. event memoryi  rate & date
Problems?
Psyco 350 Lec #18– Slide 3
Autobiographical Memory: Methods
Cue-word Method
Problems:
•
verifying event
•
dating accuracy
•
subjectivity of ratings
Psyco 350 Lec #18– Slide 4
Autobiographical Memory: Methods
Diary Studies
• Diary Phase: Participants record (and rate)
events soon after they happen.
• Test Phase: recall, cued-recall, recognition,
dating rating.
Problems?
Psyco 350 Lec #18– Slide 5
Autobiographical Memory: Methods
Diary Studies
Problems:
• restrictions on participants & events
• generalizing from diarist to non-diarist
• generalizing from recorded events to nonrecorded events.
Psyco 350 Lec #18– Slide 6
“The” Diary Study: Wagenaar (1986)
Diary Phase:
• each day for 5
events:
YEARS: record 1 or more
Test Phase (following 5-yr diary period):
• 5 events/day
• cued recall:
1 cue  2 cues  3 cues  critical detail
Psyco 350 Lec #18– Slide 7
Wagenaar (1986): Event Coding
For each event, specify:
• who
• what
• where
• when
• critical detail
For each event, rate:
• salience
• involvement
• pleasantness
Psyco 350 Lec #18– Slide 8
Wagenaar (1986): Results
Event Age affects event memory
• cued recall:  w/ age
• cued recall still well above chance after 5 years
Psyco 350 Lec #18– Slide 9
Wagenaar (1986): Results
Affect affects event memory:
• recall  w/ emotional involvement
• recall  w/ pleasantness
• unpleasant memories poorly recalled at first.
Psyco 350 Lec #18– Slide 10
Wagenaar (1986): Results
Distinctiveness affects event memory:
• cued recall  salience
Psyco 350 Lec #18– Slide 11
AM: Retention Factors
• Event Age
• Level of Affect
• Distinctiveness
• Long-term importance
Psyco 350 Lec #18– Slide 12
Retention Factors
Event Age: more recent events better recalled
than older events.
• Reasons: decay, interference, retrieval failure,
consolidation failure
Psyco 350 Lec #18– Slide 13
Retention Factors
Level of Affect: events that elicit strong
emotional response better recalled than those
that do not.
• affect related to vividness of initial encoding,
rehearsal/importance
Psyco 350 Lec #18– Slide 14
Retention Factors
Distinctiveness: distinctive/unique events tend
to be remember better than mundane/repetitive
events
• Reasons:
– mundane events tend to be schematized
– for unique events, content-based retrieval cue
accesses only one event memory.
Psyco 350 Lec #18– Slide 15
Retention Factors
Long-term Importance: important events tend to
be better recalled than unimportant events.
• Reasons:
– Importance related to level of overt/covert
rehearsal
– Important event have more “structural” support –
elaboration, organization
Psyco 350 Lec #18– Slide 16
Organization of AM
Organization of AM
Why understanding organization is important?
– directs/facilities search/retrieval through AM
– reflects encoding/post-encoding processing
Psyco 350 Lec #18– Slide 17
Hierarchical Organization
Life Periods
• temporally limited, thematically defined, concurrent
• High school years, Turkey period, PhD years…
Event Clusters/General Events
• personal narratives, mini-histories, event sequences,
generic/repeated events
• My trip to Japan, my Candidacy Exam story,
breakfasts…
Individual Events
Note: many levels of nesting possible:
subsub-eventsubeventevent
Psyco 350 Lec #18– Slide 18
Conway’s Model
Psyco 350 Lec #18– Slide 19
Conway’s Self Memory System (SMS) Model
– Life-time period:
– represents general knowledge about
characteristics of a period
– when I was at university, high school.
Psyco 350 Lec #18– Slide 20
Conway’s Self Memory System (SMS) Model
– General events:
– associated events linked by a common theme
(e.g., my trip to Toronto), or repeated events
(my breakfasts).
Psyco 350 Lec #18– Slide 21
Conway’s Self Memory System (SMS) Model
- Event-specific knowledge:
- sensory-perceptual episodic memories
(e.g.,car accident when returning from
Toronto).
Psyco 350 Lec #18– Slide 22
Conway’s SMS Model:An Example of a
Retrieval Process
• Cue-word: Computer
– “Thinking about the times when I was at the
university” (life-time period)
– “I am thinking of my first year IntroPsych
classes (general events)
– “the computer was broken in the middle of
the experiment I took part last week (ESK)
Psyco 350 Lec #18– Slide 23
An Alternative Organization (Schank; Reiser)
Activity-based
Organization
• Event memories
associated w/ activity
concepts they embody.
• Retrieval:
– start with event-type
and specify additional
features.
Psyco 350 Lec #18– Slide 24
Horizontal Organization
Horizontal organization = event-to-event links
Horizontal links, often assumed, but little studied.
Issues:
• Do event-to-event links exist?
• How are events linked horizontally?
• How are links created/maintained?
Psyco 350 Lec #18– Slide 25
Horizontal Organization: Two General Positions
Special Narrative Processing Position
• created ONLY by narrative processing given to
important life stories
Matter-of-Course Position
• normal memory processing
• higher-level cognitive processing
– planning, evaluation, comprehension
• narration
Psyco 350 Lec #18– Slide 26
Studying Event-to-event Organization
Existing Methods
Word & phase cueing and diaries
• fundamental problem: data for single events only
“tell me a story” method
• fundamental problems: selection bias, schema-driven
reconstruction.
Psyco 350 Lec #18– Slide 27
Studying Event-to-event Organization
Nonetheless, intuition, tell-me-a-story studies
indicate:
• Event memories often part of larger narratives
Issues:
• Determine “event cluster” prevalence
• Understand clustering process(es)
Psyco 350 Lec #18– Slide 28
Studying E-to-E Organization: Event-cueing
Brown & Schopflocher (1998)
General Method:
Cueing event: auto event1 [E1]

Cued event:
auto event2 [E2]
Assumption:
E1 & E2 often associated
Implication:
pattern of E1  E2 relations reflects underlying
organization
Psyco 350 Lec #18– Slide 29
Brown & Schopflocher (1998)
Two groups (during Phase 1 only):
word-cued group
important-event group
Five Phases:
1. generate cueing events
2. event-cueing task
3. relations-coding task
4. event-dating task
5. Importance-rating task
Psyco 350 Lec #18– Slide 30
Phase 1
Important-Events Condition
• E1 acquisition: Prompt
event1 [E1]
• task: to recall important personal event
• (e.g., my mom and telling me that my dad was
going to move out)
• 14 trials
• Retrieval time was measured
• Memories (E1s) were typed
• 72 undergrads
Psyco 350 Lec #18– Slide 31
Phase 1
Word-Cue Condition
• E1 acquisition: Cue word
event1 [E1]
• task: to recall a personal event related to the
cue word
• CAR
when I was 15, I stole a car.
• 14 nouns
• Retrieval time was measured
• Memories (E1s) were typed
• 73 undergrads
Psyco 350 Lec #18– Slide 32
Phase 2
Event-cueing Task
• Identical for both important-events and wordcue conditions
• E2 acquisition: event1 [E1]
event2 [E2]
• My mom and dad telling me my dad was going
to move out [E1]
Dad leaving a note on my
pillow saying bye [E2] .
• All E1s served as cues
• Retrieval time was measured
• Memories (E2s) were typed
Psyco 350 Lec #18– Slide 33
Sample Responses: Word-Cued Condition
4 2
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1 1
letter this reminds me of the first letter I wrote to my best friend in Toronto
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
10
10
11
11
12
12
13
13
14
14
letter
book
book
flower
flower
game
game
dollar
dollar
river
river
machine
machine
dog
dog
picture
picture
car
car
hand
hand
window
window
tree
tree
box
box
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
When she wrote back telling me that she was returning to Edmonton
When I bought my mother flowers for one of her birthdays
I remember the huge hug I received in return from my mother
The first game I ever received was Monopoly on Christmas
I remember winning the first time I played it against my brother.
I think of the 20 dollar bill I found on my way to school
I remember going to the store and buying a month's supply of candy (grade 4)
I think of the barbeque my family had beside the river in China in 1993
I remember barbecuing fish by the river which tasted okay.
when my dad bought me my first typewriter
I remember when I kept everyone awake while typing in the middle of the night
When I got bitten by a dog on my way to school in grade 1
I remember that day was the first time I was absent for school
I think of the time my family got our portraits taken in 88
I remember in 3/4 of the proofs, my eyes were either half or fully closed
When I finally got my first car on my 18th birthday
I remember emptying out my bank account to pay for half of the cost.
When I glued my hands together during my first experience with Crazy Glue
I was trying to glue back my mother's statue that I accidently broke
When my thumb got smashed while closing the window very quickly
I remember my thumbnail turning a hideous black color and falling off.
When I crashed into a tree during tobogganing with my friend
I remember the huge ugly bruises on my legs afterwards
I think about the first jewerly box I got from aunt on 7th birthday
I remember the necklace she gave me along with the box
Psyco 350 Lec #18– Slide 34
Phase 3
Relation-Coding Task
• Identical for both important-events and wordcue conditions
• [E1 & E2] + Relation Menu
Select relation(s)
• All event pairs scored
• Unrestricted selection
• Untimed
Psyco 350 Lec #18– Slide 35
Phase 3
Relation-Coding Task
• Event A: My mom and dad telling me my dad was
going to move out.
• Event B: Dad leaving a note on my pillow saying bye
–
–
–
–
–
–
–
Did event A and B involve the same person or persons?
Did event A and B involve the same activity?
Did event A and B involve the same location?
Did one of the events cause the other?
Is one of the events part of the other?
Are both events part of a single broader event?
Are event A and event B related in some other way?
Psyco 350 Lec #18– Slide 36
Phase 4
Event Dating Task
• Identical for both conditions
• Ex
Date for Ex
• My mom and dad was telling me my dad was
going to move out
June 30, 1986
• All events dated
• Random presentation
• Untimed
Psyco 350 Lec #18– Slide 37
Phase 5
Importance Rating Task
• Identical for both conditions
• Ex
Importance rating for Ex (1-to-5 scale)
• My mom and dad was telling me my dad was
going to move out
5
• All events rated
• Random presentation
• Untimed
Psyco 350 Lec #18– Slide 38
Horizontal Organization: Two General Positions
Special Narrative Processing Position
• created ONLY by narrative processing given to
important life stories
Matter-of-Course Position
• normal memory processing
• higher-level cognitive processing
– planning, evaluation, comprehension
• narration
Psyco 350 Lec #18– Slide 39
Competing Predictions
Narrative Position:
• Only important events will frequently cue
cluster mates.
Matter-of-course Position:
• Important and unimportant events will
frequently cue cluster mates
Psyco 350 Lec #18– Slide 40
B&S Results: Event Age
Psyco 350 Lec #18– Slide 41
B&S: Results
Defining clustered Pairs:
• Clustered Pairs:
– Either cause, subevent, or "same story" relation
indicated.
• Nonclustered Pairs:
– Neither cause, subevent, nor "same story" relation
indicated.
% Clustered:
• Important-Events Group:
• Word-Cued Group:
Psyco 350 Lec #18– Slide 42
81%
75%
Percentage of Clustered Pairs
B&S Results: Clustering & Importance
100
90
80
70
60
50
40
30
20
Important-Events Group
Word-Cued Group
10
0
1
2
3
4
5
Cueing Event Importance
Psyco 350 Lec #18– Slide 43
B&S Results: Clustered vs Nonclustered Pairs
Median Time(sec) to Retrieve Cued event (E2)
importantevent
word-cued
6.8
6.6
nonclustered 8.4
8.2
clustered
Median Difference(Days) between Cueing (E1)
& Cued Event (E2)
importantevent
word-cued
clustered
2.0
1.0
nonclustered
210.5
319.5
Psyco 350 Lec #18– Slide 44
B&S Results: Clustered vs Nonclustered Pairs
Interevent Relations as a Function of Clustering
Same
Person
Same
Location
Same
Activity
Other
54%
52%
39%
16%
nonclustered 44%
33%
32%
32%
clustered
Psyco 350 Lec #18– Slide 45
Summary
Main Claims:
• Event clusters very common
• Clustered events:
– causally & thematically related
– temporally proximate
Evidence:
• 80% clustered
• RT: clustered < nonclustered
• Age s: clustered << nonclustered
• Overlapping story elements:
– clustered > nonclustered
Psyco 350 Lec #18– Slide 46
Interpreting Event Clusters
• A Strong Narrative Position
• Event clusters are narratives.
• Narrative processes necessary for creation/
maintenance of autobiographical memory.
Psyco 350 Lec #18– Slide 47
Interpreting Event Clusters
• A Comprise Position
• Clustering prompted by:
– temporal contiguity
– similarity
– causal reasoning
– goal directed planning and evaluation
• Narrative creation/maintenance facilitated by
clustering
Psyco 350 Lec #18– Slide 48
Retrieval Processes in Autobiographical
Memory (Uzer, Lee & Brown, 2011)
Models of Autobiographical memory (AM)
assume two retrieval strategies:
– Generative retrieval
– Direct retrieval
Generative retrieval has been assumed as
a normative form of retrieval
Psyco 350 Lec #18– Slide 49
Introduction
Retrieval processes: Reaction time (RT)
Studies have shown:
– Concrete nouns < emotion terms
– Personal periods < common activities
– Common activities < general actions
Psyco 350 Lec #18– Slide 50
Introduction
RT differences:
– As an index of effort required to generate
appropriate cues
– Generation is easy and fast, when cue
accesses to associative links
– Generation is slower and harder, when cue
must be reformulated to access to
associative links
Psyco 350 Lec #18– Slide 51
Introduction
Dual Retrieval Strategies Approach:
– RT differences reflect multiple retrieval
strategies
– RTs reflect weighted blend of generative
and direct retrieval modes
Psyco 350 Lec #18– Slide 52
Current Questions
• How common are direct and generative
retrieval?
• Would retrieval strategy change as a
function of task conditions (e.g., the
types of cues provided)?
• Would RT differences still occur holding
retrieval strategy constant?
Psyco 350 Lec #18– Slide 53
Exp 1: Method
40 U of A undergraduate students
Within-subjects design
Procedure: Cue word :
“specific
personal event that is related to the cue
word”
Two types of cue words:
– Concrete nouns (e.g., chair, pencil, book)
– Emotion terms (e.g., shy, happy, sad)
Psyco 350 Lec #18– Slide 54
Exp 1: Method
Concurrent verbal protocols:“Think out loud
by verbalizing all your thoughts as you are
thinking them”
Pressing SPACEBAR as memory is retrieved
(RT)
Strategy Report: Participants were asked if
the memory came directly into mind - “Y” for
YES or “N” for NO
Brief written report of memory
Psyco 350 Lec #18– Slide 55
Exp 1: Results
Psyco 350 Lec #18– Slide 56
Exp 1: Method
Psyco 350 Lec #18– Slide 57
Exp 1: Method
Psyco 350 Lec #18– Slide 58
Exp 1: Method
Psyco 350 Lec #18– Slide 59
Exp 1: Issues
Concurrent verbal protocols: Reactivity to
generative retrieval
“Did this memory came immediately to mind”
might create a demand for directly retrieved
memories.
Psyco 350 Lec #18– Slide 60
Exp 2: Method
300 U of A undergraduate students
Procedure: Cue-words ---- AM (Identical to
Exp 1.)
Pressing SPACEBAR as memory is retrieved
(RT)
Psyco 350 Lec #18– Slide 61
Exp 2: Method
Strategy Report:
Cond. 1 (Direct): “Did the memory come
immediately to mind” Yes or No responses
Cond. 2 (Generative ) : “Did you actively
search to recall this memory?” Yes or No
responses
Cond. 3 (Both Option): “How did you retrieve
this memory?”
-The memory came immediately to mind
-I actively searched to recall the memory
Psyco 350 Lec #18– Slide 62
Generative Ret.
Direct Ret.
60
55
50
45
40
35
30
25
20
15
10
5
0
60
55
50
45
40
35
30
25
20
15
10
5
0
Concrete
Psyco 350 Lec #18– Slide 63
Emotion
Concrete
Emotion
Direct Condition
Generative Condition
12
11
10
Median RTs (sec)
9
8
7
6
5
4
3
2
1
0
Direct
Psyco 350 Lec #18– Slide 64
Generative
Direct Condition
Generative Condition
8
7
Median RTs (sec)
6
5
4
3
2
1
0
Concrete
Psyco 350 Lec #18– Slide 65
Emotion
Generative Ret.
Direct Ret.
13
12
11
10
9
8
7
6
5
4
3
2
1
0
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Concrete
Psyco 350 Lec #18– Slide 66
Emotion
Concrete
Emotion
Exp 2: Issues
Problems:
Participants might be deciding about their
retrieval strategies based on time.
Asking people whether memory directly came
to mind or you searched is confounded with
time.
When you directly retrieve it is fast when you
search it takes time.
Psyco 350 Lec #18– Slide 67
Exp 3: Rationale
Another approach: looking at “use of
information”.
Direct Retrieval: You don’t use any
information to find the memory, the
memory comes straight to your mind.
Generative Retrieval: To find the memory you
have to search for/use some type of
information (e.g., people in your life, any
time period, places you have been, etc).
Psyco 350 Lec #18– Slide 68
Exp 3: Method
300 U of A undergraduate students
Procedure: Cue-words ---- AM (Identical to
Exp 1.)
Pressing SPACEBAR as memory is retrieved
(RT)
Psyco 350 Lec #18– Slide 69
Exp 3: Method
Strategy Report:
Condition 1 (Direct Cond.):“This memory was
triggered by the cue word so I did not have
to use information about my life to help me
recall this memory.” Yes or No responses
Yes: Direct Retrieval
Psyco 350 Lec #18– Slide 70
Exp 3: Method
Strategy Report:
Condition 2 (Generative Cond.) : “This
memory wasn’t triggered by the cue word so
I had to use information about my life to help
me recall this memory.” Yes or No responses
Yes: Generative Retrieval
Psyco 350 Lec #18– Slide 71
Generative Ret.
Direct Ret.
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
Concrete
Psyco 350 Lec #18– Slide 72
Emotion
Concrete
Emotion
Direct Condition
Generative Condition
8
7
Median RTs (sec)
6
5
4
3
2
1
0
Concrete
Psyco 350 Lec #18– Slide 73
Emotion
Direct Condition
Generative Condition
12
11
10
Median RTs (sec)
9
8
7
6
5
4
3
2
1
0
Direct
Psyco 350 Lec #18– Slide 74
Generative
Direct Ret.
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Generative Ret.
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Concrete
Psyco 350 Lec #18– Slide 75
Emotion
Concrete
Emotion
Significance and Future Questions
Showed the prevalence of direct retrieval in
autobiographical memory
Utility of using retrieval strategy and RT
measures to interpret RT differences in
retrieving autobiographical memories,
Accounted for cue type effect in
autobiographical memory retrieval
Psyco 350 Lec #18– Slide 76