Memory_Ch8_all - ameipc - home

Download Report

Transcript Memory_Ch8_all - ameipc - home

Memory and Cognition
PSY 324
Topic 7: Everyday Memory & Memory Errors
Dr. Ellen Campana
Arizona State University
Intro to Everyday Memory


So far we’ve talked about a lot of different types
of long-termmemory, but many of the studies
have been about memory for lists of words
In this chapter we’ll be talking more about how
long-term memory functions in everyday life
Autobiographical memory
 Flashbulb memory
 Memory as a constructive process


Source memory
Autobiographical
Memory
Autobiographical Memory

Autobiographical memory = recollection of
events that belong to a person’s past

Mostly episodic memory

Field perspective – as if experiencing the event


Observer perspective – as if seeing the event (seeing self)


Common for more remote memories
Also includes semantic memories related to the self


Common for recent events
Where you were born, what your first word was, etc.
Much more complex than memory of word lists!
Autobiographical Memory

Autobiographical memory is multidimensional


Spatial, emotional, sensory, etc.
Visual information plays a large role

Greenberg & Rubin (2003) – neurophysiological evidence
(visual cortex damage)


Can’t recognize objects => Autobiographical memory impaired, even
non-visual aspects
Autobiographical memory is much richer than
laboratory memory

Cabeza & Coworkers (2004) – brain imaging study
Richness of Autobiographical Mem.

Cabeza & Coworkers (2004)

Gave participants cameras to take pics of 40 specific
campus locations
A-photos: photos individuals took themselves
 L-photos: lab photos (taken by another student)

Richness of Autobiographical Mem.

Cabeza & Coworkers (2004)

Gave participants cameras to take pics of 40 specific
campus locations
A-photos: photos individuals took themselves
 L-photos: lab photos (taken by another student)


Phase 1: exposed to A-photos & L-photos


Not in the scanner
Phase 2: shown A-photos & L-photos (some new,
some from before)… had to say which was which

This part was in the scanner
Richness of Autobiographical Mem.

Cabeza & Coworkers (2004): Findings

A-photos & L-photos activated many of the same
structures
Medial Temporal Lobe: episodic memory
 Parietal Cortex: scene processing


A-photos activated many additional areas
Regions that process memory for self
 Regions that process memory for visual space
 Regions associated with experience of mental time travel
 Hippocampus

Richness of Autobiographical Mem.

Why talk about that imaging study?

Demonstrates that autobiographical memory is
richer than lab memory
Even though participants were viewing the same location
for A-photos and L-photos, A-photos activated more
areas than L-photos.
 Activation thought to be related to memories of the
experience of taking the picture


If even simple, relatively unimportant memories
have rich representations, what makes some
“stand out” in our minds more than others?
Memory Over the
Lifespan
Memory over the Lifespan

What memories tend to “stand out”?
Personal milestones
 Highly emotional events
 Events that become big parts of a person’s life
 Major transition points

Juniors & seniors recalled more events from Sept of
freshman year than any other month
 Alumni recalled both September of freshman year and
end of senior year events more than other times

Memory over the Lifespan

Memory is different for different times in life
Memory between ages 10 and 30 “stands out”
 This effect is called the reminiscence bump


See figure 8.3 to see what this looks like on the
graph
Memory over the Lifetime

Why the reminiscence bump?

Life-narrative hypothesis
People tend to assume their life identities between 10-30
 Many important “firsts” between 10 and 30


Cognitive hypothesis
Encoding is better for periods of rapid change, followed
by stability (10-30=change, stability after that)
 Evidence: bump is shifted for late vs. early immigrants

Memory over the Lifetime

Why the reminiscence bump?

Life-narrative hypothesis
People tend to assume their life identities between 10-30
 Many important “firsts” between 10 and 30


Cognitive hypothesis
Encoding is better for periods of rapid change, followed
by stability (10-30=change, stability after that)
 Evidence: bump is shifted for late vs. early immigrants


Cultural script hypothesis

Cultural expectations shape recall (typical events, times)
Memories Across the Lifespan


All three hypotheses probably contribute to the
occurrence of the reminisence bump
These also interact with the other factors that
make certain memories “stand out”
Milestones, transitions, events that are part of our
major “life story”
 Highly emotional events


Next: flashbulb memories (emotional events)
Flashbulb Memory

Flashbulb memories are episodic memories
about the context in which you found out about
highly emotional events
Often culturally-relevant events (JFK, challenger,
911)
 Remembered for a long time, in vivid detail


Why called “flashbulb”? Brown and Kulik used
this analogy in a paper about JFK’s assassination
Their paper looked at memory after many years…
 … but were these memories accurate?

Flashbulb Memory

To test accuracy of flashbulb memories,
researchers use a method called repeated recall
Right after an event, people write about it (baseline)
 Later they are asked to write about it again
 Researchers compare writings for each person


Neisser & Harsch (1992) – repeated recall for
the challenger explosion
Neisser & Harsch (1992)
Quote from day after event:
I was in my religion class and some people walked in and
started talking about [it]. I didn’t know any details
except that it had exploded and the schoolteacher’s
students had all been watching, which I thought was so
sad. Then after class I went to my room and watched the
TV program talking about it, and I got all the details
from that.
Neisser & Harsch (1992)
Same person, quote from 2 ½ years later:
When I first heard about the explosion I was sitting in my
freshman dorm room with my roommate, and we were
watching TV. It came on a news flash, and we were both
totally shocked. I was really upset, and I went upstairs to
talk to a friend of mine, and then I called my parents.
Schmolke and coworkers (2000)

O.J. Simpson trial
Response at 3 days: I was in the commuter lounge at
college and saw it on TV. As 10:00 approached,
more and more people came into the room…
 Response at 32 months (same person): I first heard it
while I was watching TV at home in my living room.
My sister and father were with me…


Conclusion: “it seems unlikely that so-called
flashbulb memories differ from ordinary
episodic memories in any fundamental way.”
Are Flashbulb Memories Special?

Talarico and Rubin (2003) – 911 event

Some evidence that they are not special


Accuracy and number of details decreased similarly for
flashbulb and everyday episodic memories
Some evidence that they are special
Participants’ beliefs that memories were accurate stayed high for
flashbulb, but dropped for everyday episodic memories
 Participants’ ratings of vividness and how well they could
“relive” events stayed high for flashbulb, but dropped for
everyday episodic memories

Are Flashbulb Memories Special?

Davidson and Coworkers (2006) – also 911

Found flashbulb memories more resistant to fading
congruence scores high for flashbulb, but dropped for
everyday events
 All participants remembered 911, but only 65% were able
to recall what the other memory was

Are Flashbulb Memories Special?

Davidson and Coworkers (2006) – also 911

Found flashbulb memories more resistant to fading
congruence scores high for flashbulb, but dropped for
everyday events
 All participants remembered 911, but only 65% were able
to recall what the other memory was


Why?
Flashbulb memories about emotional events (amygdala)
 Added rehearsal: narrative rehearsal hypothesis

Are Flashbulb Memories Special?

What’s the point? Are they or aren’t they?

Two studies found conflicting results

Differed with respect to cues




Talarico & Rubins – participants created their own… therefore
remembered everyday memories better (Mantyla from last time)
Robinson and Coworkers – participants used given cues
Still an open debate
About memory in general, this tells us…
Emotional context of event can influence memory
 Knowledge (even if it comes later) can affect the
original memory

The Constructive
Nature of Memory
The Constructive Nature of Memory

People reporting memories unknowingly
Omit details
 Distort or change things that actually happened
 Report things that never actually happened


Constructive approach to memory

What people report as memories are constructed by
the person based on what happened plus additional
factors (knowledge, experience, expectations)
“War of the Ghosts”
Bartlett (1932)


Description of myth from an unfamiliar culture
Repeated production – participants came back
a number of times to tell the same story


Similar to repeated recall for flashbulb memories
Participants made more errors over time

Errors reflected something about the process
Story changed to be consistent with participants’ culture
 Confusing details left out
 Details changed (canoe -> boat)

Educated Guesses

Bahrick and Coworkers (1996)

Recall your own high school grades



Why these types of errors?


People tend to remember positive events more
Memory is constructive


A recall = 89%
D recall = 29%
If they were A/B students, they make a prediction that a specific
grade was most likely an A, and remember it that way
How does constructive memory work?
Source Monitoring

Source monitoring (and errors) is part of the
explanation for how constructive memory works

Source memory is the memory for how you acquired a
certain memory



Did I hear that on the news or did someone tell me?
Source monitoring errors / source misattributions
are when you remember the fact, but think it came
from the wrong source
Memories disconnected from source over time
Source Monitoring Errors

Jacoby and Coworkers (1989) – becoming
famous overnight
Acquistion: read a list of names (all made up, and
people were told this just after reading)
 Immediate test: identify famous names from list

Non-famous, new names
 Names from prior list
 Famous names


Delayed test: same as immediate, 1 day later
Source Monitoring Errors

Jacoby and Coworkers (1989) – becoming
famous overnight
Results: made-up names from the list that people
saw in the acquisition stage were more likely to be
rated as famous (in participants’ memory these
fictitious people “became famous overnight”)
 Explanation: Made-up names more familiar than
new names during delayed test, but participants
didn’t remember why they seemed so familiar
(source), so they attributed familiarity to famousness

Constructive Memory

Source misattribution involves participants
making inferences about the source, and using
those to construct / reconstruct memory
Flashbulb memory studies
 Bartlett’s war of the ghosts experiment
 Bahrick & coworkers’ high school grades experiment
 Jacoby & coworkers’ becoming famous overnight

Memory and Inference

So memory as a constructive process
What we remember is constructed based on events,
percepts, experience and knowledge
 The process of memory construction involves
making inferences
 This is a normal consequence of a largely adaptive
memory– in other words “it’s a good thing.”

Brewer (1977) /
McDermott & Chan (2006)





The children’s snowman vanished when the
temperature reached 80.
The flimsy shelf weakened under the weight of
the books.
The absent-minded professor didn’t have his car
keys.
The karate champion hit the cinder block.
The new baby stayed awake all night.
…… people who read these sentence often
inaccurate remembered the sentences as…
Brewer (1977) /
McDermott & Chan (2006)





The children’s snowman vanished
melted when the
temperature reached 80.
collapsed under the weight of
The flimsy shelf weakened
the books.
losthave his car
The absent-minded professor didn’t
keys.
The karate champion broke
hit the cinder block.
The new baby stayed
awake all night.
cried
Pragmatic Inference

Changes in wording we just saw are examples of
pragmatic inference
We use our experience to fill in the details without
even realizing we are doing it
 Two more in the book you should know about

Bransford & Johnson (1973) – hammer example
 Arkes & Freedman (1984) – baseball example


Inference uses knowledge and experience, which
are represented by schemas and scripts
Schemas

Schemas are representations of what is usually
or prototypically involved in an experience
A kids’ birthday party, a movie, oktoberfest, holidays
 Even if a movie theater is out of popcorn when you
go, you may later infer there was popcorn there
 Brewer & Treyens (1981) - office study


This type of representation is what causes us to
substitute “cried” for “stayed up all night” in the
sentences we heard earlier
Scripts

Scripts are representations of the sequence of
actions that usually occur during an experience
Restaurant, class, sports, evening out, holidays
 Even if you may not have gone trick-or-treating one
Halloween, you may remember doing so later
 Bower and Coworkers (1979) – dentist office study

Bower and Coworkers (1979)
The Dentist
Bill had a bad toothache. It seemed like forever
before he finally arrived at the dentist office. Bill
looked around at the various dental posters on
the wall. Finally the dental hygienist checked and
x-rayed his teeth. He wondered what the dentist
was doing. The dentist said that Bill had a lot of
cavities. As soon as he’d made another
appointment, he left the dentist’s office.
Bower and Coworkers (1979)


Study included many stories like “the dentist”
After a delay, participants were given titles and
asked to write what they remembered about the
stories
“Bill checked in with the dentist’s receptionist” ….
 ….Common even though it wasn’t really there


People filled in details based on their own script
for going to the dentist
Construction: Plusses & Minuses

Inferences in memory are much like gestalt laws
in perception – they are also heuristics

Heuristics – provide a “best guess” which is fast,
efficient and easy to store, but sometimes inaccurate


Hammer example – how inefficient it would be to have to
make all these details clear every time!
Algorithms – opposite of heuristics, provide an
accurate answer but take time, capacity, etc.
In the case of memory, require infinite storage
 S in the book was said to have this (it wasn’t all good!)

Construction: Plusses & Minuses


Like gestalt laws, in practical terms our
inferences are very often right
Sometimes our memories can be manipulated by
suggestions by others
Memories created
 Memories changed


Ramifications for law and the court system
Memory and
Suggestion
The Power of Suggestion

People are suggestible
Advertisements, political arguments affect our
attitudes, beliefs, and behaviors
 Information presented by others can also influence
our memory for past events


Misinformation Effect – misleading
information presented after a person witnesses
an event can change how the person describes
the event later

Misleading information in this situation is called
misleading postevent information (MPI)
Method: Presenting MPI

Step 1: Information to be remembered is
presented


Step 2: MPI is presented to some participants
and not others


Could be a list of words, film, slideshow, etc.
MPI is presented in a natural way so participants do
not know they are being misled
Step 3: All participants report on their memory
of the information presented in step 1

Reports compared for the two groups of participants
Loftus & Coworkers (1978)

Step 1: slide show of accident


Step 2: participants answer questions



Car stops at a stop sign, then turns corner and hits pedestrian
MPI group: “did another car pass the red Datsun while it was
stopped at the yield sign?”
Control:“did another car pass the red Datsun while it was
stopped at the stop sign?”
Step 3: participants see pictures, say which are old


Critical picture: car parked at a yield sign (new picture)
MPI group more likely to say yes (which was an error)
Loftus & Coworkers (1978)


Study demonstrates that misleading postevent
information (MPI) can affect details of what
people remember seeing
Another similar study looks at how MPI can
affect the conclusions people draw about other
aspects of the situation

Also replicates the finding from Loftus &
Coworkers (1978)
Loftus and Palmer (1974)


Participants watched a video of a car crash
Participants answered questions about the film
MPI group: How fast were the cars going when they
smashed into each other?
 Control: How fast were the cars going when they hit
each other?


Answers compared
MPI group (smashed) answered 41 mph, on avg.
 Control group (hit) answered 34 mph, on avg.

Loftus (1993, 1998)

Participants from Loftus & Palmer invited back
one year later
Question: Did you see any broken glass? (no = right)
 Results

MPI group (smashed) answered yes 32%
 Control group (hit) answered yes 14%


More evidence that MPI affects memory for
details (even details that are inferred)

How does this happen???
What Causes the
Misinformation Effect?

Memory-trace replacement hypothesis
(Loftus)
MPI replaces memories for the original event
 Process of consolidation is a possible mechanism


Retroactive interference hypothesis


Recent information interferes with (but doesn’t
replace) previously learned information
Source monitoring error hypothsis

Both memories are stored, but during recall we
forget which came from which source
Evidence: Source Monitoring Error

Lindsay (1990)
Day 1: People saw slides, narrated by female voice
 Day 2: People heard story (no slides) with a few
details changed (brand names, etc.)

All MPI, no control group
 Some participants heard a male voice tell the story, while
others heard a female voice tell the story



Data: answers to questions about details that were
changed for the story and details that were not
Why are voices important? Voice can be a cue to
source, which may make it easier to remember
Lindsay (1990)
Misled details
Control details
% “suggested”
responses
(wrong)
Female
Voice
Male
Voice
Source Monitoring Errors


False memories for early events also related to
source monitoring
Hyman, Jr. and Coworkers (1995)
Got stories from parents of college students
 Asked students about these events (and fake ones)
 With repeated recall, false memories were induced


How is this a source monitoring error?

Familiarity caused people to believe it happened, and
then they happily filled in details via inference
More on False Memories for Early
Events

Lindsay and Coworkers (2004) – Slime study


False memory effect stronger with a picture
DuBreuil and Coworkers (1998) – Mobiles
False memories can come out during hypnosis
 People are very confident that these are real

Eyewitness Testimony
Memory and Law

Eyewitness testimony – when someone who
was present at a crime reports about what he or
she saw
Most convincing types of evidence for a jury
 Witness confident-> evidence even more convincing



Accuracy & Confidence correlation is 0.29
Errors have of eyewitness testimony have
resulted in the conviction of innocent people
Memory and Law


David Webb, Charles Clark, Lenell Gertner all
went to jail (later released when testimony was
discovered to be inaccurate)
40 cases of exoneration based on DNA evidence
that became available after conviction

36 involved erroneous eyewitness testimony
8.5 average years in prison
 5 death sentences

Perceptual Errors


Eyewitness testimony can be affected by both
perception and memory – we’ll start with
perception
Two studies presented videos of crimes to
participants and then gave them photos, asked
them to identify the perpetrator


100% / 61% picked someone from the photos, even
though the correct person wasn’t in the photos
In actual crime scenes, many causes…
Crime Scenes

Errors associated with Attention

Easterbrook (1959) – as arousal increases, attention
narrows


crime scenes arousal is high, relevant details can be missed
Weapons focus – presence of a weapon narrows
attention

Worse if weapon is fired (Stanny and Johnson, 2000)

Decreases memory for victim, perpetrator AND weapon
Crime Scenes

Errors due to familiarity
Crime scenes involve perpetrator, victim, and
innocent bystanders
 Bystanders can be incorrectly identified as
perpetrators based on familiarity

Memory researcher Donald Thompson
 Sailor example


Confirmed in lab studies
Crime Scenes

Errors due to suggestion
“did you see the white car?” – later may have a false
memory of a white car
 “which one of these men did it?” – implies that the
person is in the lineup



Witness selects one, then becomes confident over time
“that one???” “OK” – confirmation of the witness’s
choice (which will inflate confidence in the choice)

Wells and Bradfield (1998) confirmed in the lab
Crime Scenes

Increasing confidence due to postevent
questioning

Shaw (1996) – participants were more confident in
false memories of items from an apartment if they
had been asked questions about them
Answering questions about an object makes it easier to
retrieve memories about it later
 People mistake ease of retrieval for accuracy

What can be done?


First, recognize the problem and communicate it
to jurors
Make some procedural changes
Inform witnesses that perp may not be in the lineup
 In lineups, use “fillers” that are similar to the suspect
 When presenting a lineup use sequential rather than
simultaneous presentation
 Improve interviewing techniques


Cognitive interview (25-60% more information than
police interview)