Transcript Chapter 12

+
Chapter 12
Memory in Childhood
+
2
Infantile Amnesia
 Infantile Amnesia:

The tendency for adults to have few autobiographical
memories from below the age of 5
 Studying


infantile amnesia can be difficult, as it is:
Hard to verify memories from childhood -- researchers focus
on dateable, verifiable events, such as the birth of a sibling
Hard to know whether childhood memories are genuine
recollections or are reconstructed from stories and
photographs
 Genuine memories tend to be more visual, less verbal,
more emotional, more complete (Crawley & Eacott, 2006)
+
3
Infantile Amnesia

Sheingold and Tenney (1982):
 Participants:
 College students and
children (ages 4–12)
 Task:
 Answer specific questions
about a sibling’s birth from
when they were 3–11 years
old
 e.g. “Who took care of
you while your mother
was in the hospital?”
 Mothers were asked the
same questions

Results:

If the birth occurred after 3
years old, very little forgetting
occurred


Even if it occurred many
years ago
If birth occurred before 3
years, virtually nothing was
remembered
+
4
Mnemonic Abilities in Infancy

Infants’ linguistic skills are highly
limited, so:

Experimenters can’t use verbal
instructions

Tasks require motor, rather
than verbal responses


But infants’ motor responses
are also limited
It is hard to establish whether
they are consciously aware of
their memories

i.e. are these declarative or
implicit memories?

Memories are considered declarative,
provided they pass two filters
(Richmond & Nelson, 2007):
 Amnesia Filter:
 If an amnesiac can do the task,
then it’s implicit.
 If not, then it’s declarative.
 Parameter Filter:
 If the memory is affected by
factors known to influence
declarative tasks in adults, then
it’s also declarative, such as:
 Changes in study time
 Retention interval
 Contextual changes
+
5
Mnemonic Abilities in Infancy
Implicit Memory in Development

Infants display some (limited) mnemonic abilities almost immediately
after birth
 DeCasper and Fifer (1980):
 Task:
 3-day-olds learned that sucking on a pacifier activated a tape
recording with the voice of either:
 The infant’s mother
 A stranger
 Results:
 Infants sucked on the pacifier more when it activated the
familiar voice of their mother
 Querleu et al. (1984) replicated the results in infants 2 hours old
 Conclusion:
 Newborns have the ability to remember both their mother’s
voice and the action that produces the sound of her voice
+
6
Mnemonic Abilities in Infancy
Declarative (Explicit) Memory Development
 Declarative
memory abilities develop later than
implicit memory.
 Originally,
infants were assumed to lack explicit
(declarative) memory.


This view has changed in light of recent data.
It is now assumed that declarative memory is possible in
children far younger than once thought.
 http://www.youtube.com/watch?v=lPJiB-oGMN0
+
7
Declarative Memory Development
Rovee-Collier’s (1989) Mobile Conjugate
Reinforcement Paradigm

Conjugate means “paired.”

Guiding Principle: Avoid
underestimating baby’s memory
by using what interests them
 e.g. a colorful mobile hanging
over a baby’s crib
 Attach it to the baby’s foot
with a ribbon
 When the baby kicks, the
mobile moves
 They are rewarded when
they kick by the mobile’s
movement
 Quickly learn to kick in the
mobile’s presence via
operant conditioning

Three phases of the paradigm:
 Baseline:
 Record how often the baby kicks
when the foot isn’t attached to the
mobile
 Learning:
 Infants learn that kicking (response)
causes the mobile to move
(reinforcement)
 Test:
 The ribbon is detached from the
mobile (no reinforcement during
test)
 If they kick more than baseline in
the presence of the mobile, they
remember the connection
 The retention interval between
learning and test can be
manipulated
+
8
Declarative Memory Development
Rovee-Collier’s (1989) Mobile Conjugate
Reinforcement Paradigm

Rovee-Collier et al.’s (1980)
results:

At short delays:


After 2 days:


Both 2 and 3-month-olds
showed evidence of
retention
2-month olds were back at
baseline
After a week:

3-month olds still show a
reliable effect

Presenting a reminder (a moving
mobile) before testing reactivated
kicking:
 After a 2-week delay: Retention
bounced back up to its initial levels
 After a 1-month delay: Still
significant kicking behavior
+
9
Declarative Memory Development
Rovee-Collier’s (1989) Mobile Conjugate
Reinforcement Paradigm

The learned kicking behavior is quite specific:

Perceptual discrimination:


If the babies were trained on a mobile with yellow blocks, they
wouldn’t respond to a mobile with metal butterflies instead

However, if they’re trained on many different mobiles, they
would then generalize the kicking response to novel mobiles

It is as if they learned the mobile “concept”
Context-sensitivity:

If an infant was trained in a crib but tested in the kitchen, they
wouldn’t kick

If the crib’s décor was changed, the amount of kicking would be
reduced
+
10
Declarative Memory Development
The Mobile Conjugate Reinforcement Paradigm

Is it declarative memory?

Probably (Rovee-Collier, 1997):

Infants’ performance is
determined by factors that are
more important in declarative
than implicit memories, e.g.:

Participant’s age

Retention interval

Context (what the mobile and
the environment looks like)

The Mobile Conjugate
Reinforcement paradigm isn’t
suitable for infants over 7
months, so:
 Hartshorn and Rovee-Collier
(1997) introduced a similar
task for older infants:
 Infants instead learn to
press a lever to make a
miniature train move
 They demonstrate memory
by pressing the lever even
when the train no longer
moves
+
11
Mnemonic Abilities in Infancy
Deferred Imitation


Background:
 Meltzoff (1985)
Meltzoff’s novel toy
 Piaget believed that infants
 Participants:
didn’t possess the ability to do
 14-month-olds
deferred imitation until
 24-month-olds
2+ years
 Exposure Conditions:
Deferred Imitation Task:
 Imitation: Observe experimenter
pull the toy apart
 Experimenter produces a
sequence with objects
 Control: Observe experimenter
move the toy in a circle
 e.g. use a mallet to hit a
metal plate
 Baseline: Give the novel toy to
the infant without pre-exposure
 Delay of varying
lengths
 Delay:
 The infant then
 Wait 24 hours before infant is
tries to imitate the
given the toy
action
+
12
Mnemonic Abilities in Infancy
Deferred Imitation

Results:

14-month-olds



45% of infants in the
experimental condition
imitated the experimenter’s
action
Only 7.5% of infants in the
baseline/control conditions
imitated
24-month-olds

70% of infants in the
experimental condition
imitated

Follow-up Experiments:

Collie and Hayne (1999)


6-month-olds remember
around 20% of the actions
they saw 24 hours earlier
Bauer et al. (2000)

60% of 16-month-olds
produced actions in the
right order after a 12-month
delay
+
13
Deferred Imitation
Evidence of Declarative Memory?
 Is
deferred imitation declarative or implicit memory?
 Evidence
that deferred imitation requires
declarative/explicit memory:


Adult amnesiacs show little evidence of deferred imitation
(McDonough et al., 1995)
 Amnesiacs are selectively impaired in declarative memory
Preverbal infants who imitated were later able to verbalize
their performance
 Only declarative memories are likely accessible to
language
+
14
Principles of Memory Development
Hayne (2004)

Older infants typically encode/store information faster than younger infants


Older infants remember information over longer delays


e.g. 18-month-olds kick in the mobile conjugate reinforcement paradigm up to 12
months after training; 6-month-olds only remember for about 2 weeks
Older infants make use of a greater variety of retrieval cues (their memories
are more flexible)


e.g. 6-month-olds require twice as much exposure to do deferred imitation,
compared to 12-month-olds
e.g. 18-month-olds still imitated when the toy was changed slightly; younger infants
did not (Hayne, Boniface, & Barr, 2000)
Forgotten memories can be retrieved when a reminder is presented

3 minutes of exposure to the experimenter moving the mobile 1 day before testing
vastly extended the retention interval during which they exhibited memory
+
15
Developmental Cognitive
Neuroscience

Numerous developments during the first 2 years of life contribute to
better mnemonic abilities:





Attention improves
Language starts to be acquired
World knowledge accumulates
The brain is developing
Schacter and Moscovitch (1984)


Implicit memory is controlled by a memory system likely present at birth
 Striatum
 Cerebellum
 Brainstem
Declarative memories depend on a late-developing memory system
 Reaches maturity between 8–10 months
+
16
Changes at the Neural Level
More myelination
More interconnections
More neurons in the
hippocampus and in
prefrontal areas
Newborn
6 Months
2 Years
+
17
Developmental Cognitive
Neuroscience

Brain regions underlying declarative memory continues to develop
well into childhood



Medial Temporal Lobe
 Hippocampus
 Dentate gyrus—develops for 1 year after birth
 Other parts not fully developed until 2–8 years
 Parahippocampal complex
Prefrontal cortex
 Synaptic density increases until 24 months
 Not fully matured until around 20 years
Axons in the central nervous system
 Continue to myelinate over the first year
 Speeds up processing and learning
+
18
Changes in Brain by Age
+
19
Sensitive Periods in Development
Neural changes correlate with sensitive periods in
development and important milestones.
+
20
Developmental Cognitive
Neuroscience
Benefits of the Approach

Does not simply describe
changes
 Helps to explain how and
why changes arise

Offers a partial
understanding of the
differences between types of
memory
 e.g. declarative vs. implicit
Limitations to the Approach

Often relies heavily on
correlations between the
rate of brain maturation and
behavioral performance
 But correlation does not
imply causation!
+
21
Further Memory Developments
Siegler (1998)

Declarative memory continues to improve into childhood, due to a
number of inter-related factors:

Basic capacity in short-term memory/working memory increases over the
years

Subvocally rehearse faster/more

Adopt better strategies

Learn and use new strategies (e.g. rehearsal)

Accumulate more knowledge


Helps form schemas to organize memories
Develop better metamemory:

Knowledge about one’s own memory and how it works

Helps children select the best strategy to use
+
22
Candid Camera Demo

CD – Remember This Message
+
23
Further Memory Developments
Basic Capacity

Gathercole et al. (2004) studied working memory (WM) development between
the ages of 4–15 years
WM Component

Test
Phonological Loop
Verbal Storage
(Digit Span)
Central Executive
Complex Memory
Span (Backward
Digit Recall)
Visuo-Spatial
Sketchpad
Visuo-Spatial
Memory (Visual
Pattern Recall)
Results:

All three components developed over time

WM structure remained fairly constant over
time
From Gathercole et al. (2004). Copyright © American Psychological Association.
Reprinted with permission.
+ Further Memory Developments
Content Knowledge

Older children possess more
knowledge


Memory performance is
often better when the
learner has a relevant
schema for new knowledge
Adapted from Chi (1978).

Results:
 Adults performed better at digit recall
 Children performed over 50% better at chess recall

Schneider et al.’s (1993) Follow-Up:
 Children and adults of equal chess expertise
performed equally
 Both performed better than chess novices

Conclusion:
 Memory for chess positions depends largely on
expertise, rather than age
Chi (1978)


Participants:

10-year-old chess
experts

Adult chess novices
Task:

Digit recall and
reproduction of chess
positions
24
+
25
Further Memory Developments
Memory Strategies

Older children are more likely
than younger ones to employ
memory strategies

Evidence comes from
categorized list recall:

Task:

Participants are randomly
presented with
words/pictures from
various categories

They are then asked to
free recall the items

Results:
 Adults (Weist, 1972)
 Rehearse words by category
 Words are recalled by category
(i.e. they “clustered”)
 Organizational strategies lead to
better recall
 Children aged 8–17
(Schneider, Knopf, & Stefanek,
2002)
 Older children used more sorting
strategies
 Older children clustered more
 Both these strategies increased
steadily over development
+
26
Older Children Use More Sorting
Strategies
(a) Free recall and (b) sorting during learning and clustering during recall assessed by ratio
of repetition (RR) at ages 8, 10, 12, and 17. From Schneider et al. (2002). Copyright ©
American Psychological Association. Reprinted with permission.
+
27
Further Memory Developments
Metamemory

Metamemory knowledge increases over development


e.g. younger children tend to drastically overestimate their memory
span (Yussen & Levy, 1975)
Metamemory knowledge is a moderate predictor of memory
performance, R =.41 (Schneider & Pressley, 1989)

A robust metamemory helps children to select appropriate learning
strategies

The correlation isn’t perfect because children might not be
motivated to use their strategies
+
28
Verbatim and Gist Memory
Brainerd and Reyna (2004)
Verbatim Memory

Contain accurate and
detailed information about
to-be-remembered stimuli
 Reflects the actual
experience
 Improves over childhood
Gist Memory

General semantic information
about to-be-remembered stimuli
 Reflects a general
understanding of an
experience
 Improves over childhood
 Children grow to extract
more meaning from
information
 e.g. they start to
spontaneously
categorize items like
adults
+
29
When Improvements Go Wrong
Brainerd and Reyna (2004)

Older children form more gist memory traces
 This is generally advantageous, but this occasionally leads
to errors when:
 The learning task leads older children to produce more
gist memories than younger children do
 The memory test requires verbatim recall/recognition
 Greater gist memory increases the likelihood of false
recall/recognition of information very similar in meaning
to the to-be-remembered information
 This has largely been studied with the Deese–Roediger–
McDermott (DRM) paradigm – false memory paradigm
+
30
Deese-Roediger-McDermott (DRM)
Paradigm



Task:
 Participants presented with a list of
words
 e.g. NURSE, SICK, HOSPITAL,
and PATIENT
 All words are related to a
missing target word (e.g.
DOCTOR)
 Participants are then asked to
recognize words they saw before,
including the missing word
Result of Interest:
 How often do people mistakenly
recall/recognize having seen the
missing target word?
Developmental Results:
 False recall/recognition increases
progressively during childhood
From Brainerd and Reyna (2004).Copyright © 2004 Elsevier.
Reproduced with permission.

Conclusion:

Older children are better at semantic
processing

So they tend to categorize all the words
under the missing word, which becomes
highly active
+
31
Declarative vs. Implicit Memory
Development

Implicit memory is pretty well established at
birth and does not seem to improve with age
 There are a few discrepant results, however.

Declarative memory starts off less developed,
but it then begins to improve rapidly.

Russo et al. (1995) compared implicit to
declarative memory development.
 Task:
 Asked children to ID degraded (or intact)
pictures of objects
 Implicit measure:
 Perceptual priming
 Declarative measure:
 Free Recall
Implicit
Declarative
Data from Russo et al. (1995).
+
32
Declarative vs. Implicit Memory
Development
 Why

the general lack of implicit improvement?
Compared to declarative memory, implicit memory involves
more basic processes.
 Implicit memory is relatively unaffected by other cognitive
skills, such as:
 WM capacity
 Content knowledge
 Strategy
 Metamemory
+
33
Autobiographical Memory in Infancy
Can infants store autobiographical memories?

Nelson (1988)




Tape-recorded the “crib talk” of Emily between the ages of 21 and 36
months
Results:
 At 21 months, Emily was recalling (often fragmented) events
 Most from the previous day
 Some from up to 6 months before!
 At 24 months, Emily was:
 Constructing explicit rules and generalizations
 Making speculations about the future
 At 36 months, Emily stopped talking to herself in the crib
Conclusion:
 2-year-olds can encode and remember specific episodes
Caveat:
 This was only a single child
+
34
Autobiographical Memory in Infancy
Can infants store autobiographical memories?

Fivush, Gray, and Fromhoff
(1987)

Interviewed ten children
(average age = 33 months)
and their parents about
recent and distant events

The children were able to
answer over 50% of
questions about both types of
events

Best recall for
activities/objects, rather
than people/location

A major factor determining how
much can be remembered:


Whether they possessed
language skills to talk about
the event at the time it
happened
At longer delays (4+ years)
before recall, the infant’s age at
encoding is highly important:

If under 2, generally no
memory

If 3, then 50% recall

If 4+, then nearly all recalled
+
35
Autobiographical Memory in Infancy
Simcock and Hayne’s (2003) “Magic Shrinking
Machine”

Memory in young children is
typically assessed by verbal
report

However, they have limited
verbal skills


Research based on verbal
report likely underestimates
young children’s memory
Simcock and Hayne developed
the “magic shrinking machine”
task to address this issue

They included nonverbal
memory tests

“Magic Shrinking Machine”
 Participants:
 Children age 24–48
months
 Task:
 Children saw large objects
go into a machine, but
small objects come out
 After 24 hours, children
were given three memory
tests:
 Verbal recall
 Nonverbal photograph
recognition test
 Behavioral reenactment
+
36
The Magic Shrinking Machine
Simcock and Hayne (2003)
From Simcock and Hayne (2003). Copyright © American Psychological Association. Reproduced with permission.

Simcock and Hayne’s nonverbal tests revealed hidden memory
retention by:

Relying less on language

Providing more retrieval cues to the children
+
37
Explanations for Infantile Amnesia
Fivush and Nelson’s (2004) Social Cultural Theory

Pre-linguistic memories are hard to express later using language.


Language skills at the time of an event dictate what they can recall
subsequently (Simcock & Hayne, 2002).
Children whose parents have an elaborative reminiscing style
later report more and fuller childhood memories.

Nelson’s (1989) Museum Study:

When mother–child conversations about the museum trip were freely
interacting, rather than practical, the children remembered more.
+
38
Explanations for Infantile Amnesia
Repression

Freud (1915/1917)


Proposed that infantile amnesia occurs through repression, with
threat-related thoughts being:

Banished to the unconscious mind

Transformed into more innocuous memories called “screen
memories”
Problems with the theory:

No evidence supports it

Fails to explain why adults cannot remember positive or neutral
events from childhood
+
39
Explanations for Infantile Amnesia
Cognitive Self Approach

Howe and Courage (1997)
 Must have a sense-of-self to form autobiographical memories.
 Develops around 2 years.
 Visual-self recognition: recognizing one’s reflection in the
mirror
 Provides a schema for autobiographical memories.
 Evidence (controlling for language):
 Self-recognizers had better memory for personal events.
 Pre-self recognizers never had good autobiographical memory.
 Why can 2-year-olds remember events for months, but not into
adulthood?
 Howe and Courage argue that these memories aren’t rehearsed
much.
+
40
Explanations for Infantile Amnesia
Fivush and Nelson’s (2004) Social Cultural Theory
 Nelson’s

(1989) Museum Study:
Elaboration provides the children with ample opportunities to
rehearse their own memories.
 An elaborative style is more common in Western Cultures.
 First memories in Western cultures tend to be earlier and
are more elaborated and emotional than in many Eastern
cultures.
 This could also be because Western children are more
inclined to talk about their personal experiences .
+
41
Explanations for Infantile Amnesia
Summary


The cognitive self approach and the social cultural theory both
have supportive evidence and are not mutually exclusive.

The onset of autobiographical memory could depend on the
emergence of self.

Subsequent memory expression is heavily influenced by social,
cultural, and linguistic factors.
Most research into infantile amnesia relies on correlational
evidence.

Causality is difficult to prove under these circumstances.
+
42
Children as Witnesses
Accuracy


Are traumatic events more memorable than
nontraumatic ones?
 Not terribly (Cordón et al., 2004):
 Both are influenced by age, delay, and
nature of the event
Are children more suggestible than adults?
 Yes (Ceci, Baker, & Bronfrenbrenner,
1988):
 Younger children are more biased than
are older children by leading
questions:
 Questions that carry with them an
implication as to the correct answer
 10- to 12-year-olds are no more
suggestible than adults
The Effect of Leading Questions
+
43
Children as Witnesses
Suggestibility

Thompson, Clarke-Stewart, and Lepore (1997) found that young children’s
responses are largely consistent with the view of their questioner.


The responses of 5 to 6-year-olds to questions about potential abuse when
questioned by:
 Neutral interviewers:
 Are generally accurate
 Accusatory interviewers:
 Are biased in favor of guilt
 Exonerating interviewers:
 Are biased in favor of innocence
Young children continue to reflect the prior influence even when:


Questioned by a new, non-suggestive interviewer.
Warned that the previous interviewer may have been mistaken.
+
44
Children as Witnesses
Suggestibility

Young children are suggestible because of their:



Social compliance
 They yield to authority figures
 They lack social support to stand up for their views
Cognitive incompetence
 They come to believe their distorted reports because of limitations
in:
 Processing
 Attention
 Language abilities
Inability to source monitor
 They often confuse real-life and television events.
+
45
Children as Witnesses
How to Maximize Accuracy

Reduce social compliance
 Avoid leading questions at any point in the questioning process
 Garven, Wood, and Malpass (2000)

Train effective source monitoring techniques
 Thierry and Spence (2002)

Reinstate the encoding context
 According to the encoding specificity principle, memory should be
maximal when the encoding context and the retrieval context match
 Priestley, Roberts, and Pipe (1999)

Use nonverbal recall techniques
 Asking children to draw what they remember before asking for a verbal
report can elicit idiosyncratic retrieval cues and nonverbal information
 Gross and Hayne (1999)
 Children remembered 30% more in the drawing condition, which
only increased (without adding errors) at longer delays