Discussion 6
nikkieramsey
6weekDefiningtheboundary.pdf
Defining the Boundary: Age-Related Changes in Childhood Amnesia
Karen Tustin and Harlene Hayne University of Otago
Childhood amnesia refers to the inability of adults to recall events that occurred during their infancy and early childhood. Although it is generally assumed that children and adolescents also experience child- hood amnesia, with limited exceptions, most empirical research on the phenomenon has focused exclusively on adults. Here, we developed a new Timeline procedure to directly compare the early memories reported by children, adolescents, and adults. Overall, the proportion of memories reported before the age of 3 years was greater for the children and adolescents relative to the adults. In addition, the single earliest memory reported by children and adolescents was at a younger age than that reported by adults. In fact, the earliest memories reported by the children and adolescents, but not the adults, were significantly younger than the traditional 3 1⁄2-year-old boundary of childhood amnesia. Regardless of the age of the rememberer, participants’ early memories had the same episodic characteristics. We conclude that the boundary and the density of childhood amnesia may increase over the course of human development and that age-related changes in basic memory mechanisms make an important contribution to our understanding of the source of childhood amnesia.
Keywords: childhood amnesia, episodic memory, children, adolescents
For more than a century, psychologists have recognized that most adults have little or no recollection of events that occurred during their infancy and early childhood. Although Freud (1905/ 1963) originally coined the terms infantile or childhood amnesia to describe this phenomenon, Henri and Henri (1898) were the very first researchers to provide empirical evidence for the paucity of memories for events that occurred early in life. In their study, 123 participants (age 16 to 65 years) from a range of countries an- swered a series of questions about their earliest recollections. Most participants had some memory of their early childhood, but there were large individual differences in the age at which these early memories occurred, ranging from 1 to 6–8 years of age. Despite this broad range, the bulk of participants’ earliest memories fell between the ages of 2 and 4 years. Most people experienced virtually complete amnesia for events that occurred during their infancy and very early childhood. Following the seminal descrip- tions of childhood amnesia by Henri and Henri (1898) and by Freud (1905/1963), researchers attempted to refine the boundaries of the phenomenon. Freud originally suggested that we experience complete amnesia for the first 6 to 8 years of life, but the vast
majority of empirical studies on the age of adults’ earliest mem- ories are more consistent with the findings originally published by Henri and Henri (see Table 1). The grand mean of the average age of earliest memory found in these studies is 3.48 years.
Empirical research on the phenomenon of childhood amnesia has now spanned over a century. There have been many advances in our understanding of the nature of the phenomenon, but there are three important limitations to the research that has been con- ducted to date. First, in the majority of studies, researchers have used adults’ single earliest memories as a proxy for the age of the offset of childhood amnesia. Although it may be true that the earliest age at which we are able to recall a specific event may give us some indication of when childhood amnesia begins to abate, it is unlikely that childhood amnesia is an all-or-none phenomenon (Newcombe, Lloyd, & Ratliff, 2007; Pillemer & White, 1989). By examining the distribution of individuals’ early memories, on the other hand, we not only can establish the age at which our first memories begin to leak through the barrier but also can examine how early memories begin to form the first chapter of our autobi- ography (K. Nelson & Fivush, 2004; Peterson, Wang, & Hou, 2009; Rubin, 2000; Wang, 2003; Wang, Conway, & Hou, 2004).
A second limitation to prior empirical research is that research- ers have yet to examine the specific episodic content of individu- als’ very early memories. The term episodic memory is used to refer to the recollection of personal, past experiences that consti- tute our autobiography. Tulving coined the term; he argued that episodic memory is memory for information about the “what,” “when,” and “where” of an event that happened to “me” (Tulving, 1972). More recently, Tulving has argued that episodic memory is accompanied by autonoetic consciousness, or an awareness that this event happened to “me” in the past, that does not accompany retrieval of other kinds of memories (see Tulving, 2002a, 2002b). Although researchers typically assume that the early memories reported by participants in studies of childhood amnesia are epi-
Karen Tustin and Harlene Hayne, Department of Psychology, Univer- sity of Otago, Dunedin, New Zealand.
This research was funded by grants from the Marsden Fund of the Royal Society of New Zealand to Harlene Hayne. Karen Tustin was supported by a New Zealand Tertiary Education Commission Top Achiever Doctoral Scholarship. We thank Fiona Jack, Debbie McLachlan, Nicola Davis, Jenny Richmond, and Michelle Tustin for their help with data collection, interview transcription, and interobserver reliability. We also thank Julien Gross, Rachel Zajac, and Gordon Harold for their advice regarding data analysis. Special thanks to all of the families who participated in our study.
Correspondence concerning this article should be addressed to Harlene Hayne, Department of Psychology, University of Otago, PO Box 56, Dunedin 9054, New Zealand. E-mail: [email protected]
Developmental Psychology © 2010 American Psychological Association 2010, Vol. 46, No. 5, 1049–1061 0012-1649/10/$12.00 DOI: 10.1037/a0020105
1049
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
sodic in nature, to the best of our knowledge, no one has actually tested this assumption. Some researchers have coded the content of early memories in children (Buckner & Fivush, 1998; Butler, Gross, & Hayne, 1995; Fivush & Schwarzmueller, 1998; Peterson, 1994; Peterson, Grant, & Boland, 2005; Van Abbema and Bauer, 2005) and in adults (Dudycha & Dudycha, 1933a, 1933b; Howes, Siegel, & Brown, 1993; Kihlstrom & Harackiewicz, 1982; MacDonald, Uesili- ana, & Hayne, 2000; McCabe, Capron, & Peterson, 1991; Mullen, 1994), but the coding schemes that have been used to evaluate the memories would not meet Tulving’s definition of episodic mem- ory, particularly in its emphasis on autonoetic consciousness. In addition, researchers have yet to directly compare the episodic content of the early memories that children and adults report.
The third limitation of prior studies of childhood amnesia is that most of the work examining the distribution and content of par- ticipants’ earliest memories has involved research with adults. Given that the most common explanations for childhood amnesia involve developmental changes that take place during childhood (for reviews, see Bauer, 2007; Hayne, 2004; Howe, 2000; K. Nelson & Fivush, 2004; Newcombe et al., 2007; Pillemer & White, 1989; Wang, 2003), studying the phenomenon in children should shed new light on the mechanisms involved.
Do children, like adults, experience childhood amnesia? In a classic study of early memories reported by children, Sheingold and Tenney (1982) interviewed 4-, 8-, and 12-year-olds about the birth of a younger sibling. They found that if the sibling had been born when the participant was younger than 4, regardless of the participant’s current age, he or she had very little recollection of the events surrounding the birth. The 4-year-old boundary obtained in the Sheingold and Tenney study was very similar to the bound- ary of childhood amnesia documented in prior studies with adults (see Table 1). These findings provided some preliminary evidence that children, like adults, might experience dense amnesia for events that occurred during early childhood and that the age of their earliest personal memory might be very similar to that of adults.
Although the Sheingold and Tenney (1982) study is highly cited as evidence of childhood amnesia during childhood, it does have some important limitations of its own. First, Sheingold and Tenney assessed participants’ recollection of a single event, the birth of a sibling. Subsequent prospective research on memory development
has shown that young children do not place the same emphasis on the birth of a sibling as do older children and adults (Fivush, Gray, & Fromhoff, 1987; K. Nelson, 1989). Thus, this particular event, for reasons other than childhood amnesia, may not be recalled by either children or adults if the birth occurred during early child- hood. Second, the participants in the Sheingold and Tenney study were asked a series of questions about their sibling’s birth. Many of these questions involved details to which children are unlikely to attend (e.g., “What time of day was it when [your mother] left to go to the hospital?”). For these reasons, the Sheingold and Tenney study may not provide an accurate estimate of the age of earliest memories in children.
In a second study conducted with children, Fitzgerald (1991) interviewed 6- and 9- to 10-year-olds about their earliest memories using a picture-cue method; this method was a derivative of the word-cue method commonly used to elicit earliest memories in studies with adults (see Crovitz, Harvey, & McKee, 1980; Crovitz & Quina-Holland, 1976; Jack & Hayne, 2007), except that pictures were used instead of words to cue participants’ memories. For example, children were shown a picture and were then asked to recall an early memory associated with that picture. Fitzgerald found that the picture-cue method was largely ineffective in elic- iting children’s early memories; in fact, children focused their attention more on naming the pictures than on recalling personal memories associated with the pictures.
In another study, Fivush and Schwarzmueller (1998) inter- viewed young children about a number of prior events on four occasions when they were between 40 and 70 months of age. They then reinterviewed these children about the same events at the age of 8 years. Fivush and Schwarzmueller found that the children recalled about 80% of the original events irrespective of the age at which the events occurred (e.g., at 40 or 70 months). In fact, during the follow-up interview, children even reported new information about the events that they had not reported during the first inter- view. More recently, Van Abbema and Bauer (2005) used a similar method to interview 3-year-old children about past events and then reinterviewed the children about the same events when they were 7, 8, or 9 years old. In contrast to Fivush and Schwarzmueller, Van Abbema and Bauer found that their participants recalled only 40% of the early events during the follow-up interview.
One possible explanation for the discrepancy in the proportion of remembered events between the Fivush and Schwarmueller (1998) and Van Abbema and Bauer (2005) studies (80% vs. 40%) is that Van Abbema and Bauer initially interviewed children about events that had taken place between 1 and 17 months prior to recall, with most of the events occurring 9 months prior to recall. Thus, during the follow-up interview, children were asked to recall events that took place when they were just over 2 years old—more than a full year younger than those in the Fivush and Schwar- zmueller study. Taken together, these two studies provide tentative support for Sheingold and Tenney’s (1982) conclusion that chil- dren, like adults, experience childhood amnesia. Because none of these studies specifically targeted participants’ earliest memories, however, the data are equivocal about potential age-related differ- ences in the boundary of the phenomenon.
In the first direct comparison of childhood amnesia in children, adolescents, and adults, Peterson et al. (2005) asked groups of 6- to 9-year-old children, 10- to 13-year-old children, 14- to 16-year- old adolescents, and 17- to 19-year-old adolescents to describe and
Table 1 Age Range, Modal Range, and Mean Age (All in Years) of Adults’ Earliest Memories
Reference Age range Modal range
Mean age
Henri & Henri (1898) �1–8 2–4 3.10 Dudycha & Dudycha (1933a) 1.5–5 3–4 3.71 Dudycha & Dudycha (1933b) 1–5 3–4 3.58 Kihlstrom & Harackiewicz (1982) — — 3.24 McCabe et al. (1991) 2–5 3–4 4.00 Howes et al. (1993) �1–10 — 3.24 Mullen (1994) — — 3.36 West & Bauer (1999) 0.7–10 — 3.33 MacDonald et al. (2000) 0–11� 3–4 3.57
Grand mean 3.48
1050 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
date their earliest self-nominated personal memory. Peterson et al. found that the age of participants’ earliest memories was influ- enced by the participants’ current age. For example, the average age of earliest memory reported by the 6- to 9-year-old children was significantly younger than was the average age of earliest memory of the 10- to 13-year-old children and of the 14- to 16-year-old adolescents. The average age of earliest memory for the 17- to 19-year-old adolescents did not, however, differ from the average age of earliest memory reported by the youngest group of children. On the basis of these results, Peterson et al. concluded that the boundary of childhood amnesia changes as a function of participants’ chronological age at the time of the interview. That is, children who are younger than 10 tend to have younger earliest memories than do children and adolescents who are older than 10.
In contrast to other studies conducted with children, Peterson et al. (2005) provides the first suggestion that younger children do not experience childhood amnesia to the same degree as do adults. It is important to note, however, that that the average age of earliest memories across all groups of children and adolescents in the Peterson et al. study was over the 3-year boundary of childhood amnesia typically found for adults. Given this, the age of the earliest memories obtained from the younger samples in that study did not differ from those commonly reported in studies of child- hood amnesia conducted with adults (see Table 1). Furthermore, the average age of earliest memories in the Peterson et al. study yielded an inverted U-shaped function; younger children reported younger earliest memories than did older children, but the older children had older earliest memories than did the oldest group of adolescents. One potential explanation for this somewhat puzzling result may have been that the youngest children did not understand the experimenter’s instructions regarding the kinds of memories they were being asked to retrieve and report. In addition, the wide age range within each age group may have obscured potential age-related differences in the age of earliest memories.
In summary, there are three important limitations inherent in prior research on childhood amnesia. First, most studies rely on the age of an individual’s earliest memory as an indication of when childhood amnesia abates (but see K. Nelson & Fivush, 2004; Peterson et al., 2009; Rubin, 2000; Wang, 2003; Wang et al., 2004). Second, there have been no detailed analyses of the specific episodic content of individuals’ earliest memories that would meet Tulving’s definition of episodic or autonoetic content. Finally, in the majority of research on childhood amnesia, adult populations have been used to study a developmental phenomenon. In the present study, we have attempted to overcome these limitations. The specific research questions were threefold: First, do children and adolescents, like adults, experience dense amnesia for events that took place during the first 3 years of life? Second, do the single earliest memories reported by children and adolescents originate from an earlier point in development than do those reported by adults? Finally, for participants of all ages, do these early recollections reflect true episodic memory? In the present experiment, we extended Peterson et al.’s (2005) study by com- paring the early and recent memories of groups of young children (5-year-olds), older children (8- to 9-year-olds), adolescents (12- to 13-year-olds), and adults (18- to 20-year-olds). The age groups that were included in the present experiment were narrower than those included in the Peterson et al. study. We also used a new procedure designed to clarify the instructions for participants of all
ages. In this procedure, each year of the participant’s life is marked on a Timeline with photographs of the participant at particular ages. In the present experiment, we used the Timeline to help children, adolescents, and adults identify their early and recent memories as well as their single earliest memory.
Method
Participants
A total of 48 children, adolescents, and adults participated in this experiment. All participants were of European descent (one child also indicated Maori ethnicity), and all were fluent English speakers. The children and adolescents were recruited from an existing pool of participants enrolled in the Early Learning Project at the University of Otago, Dunedin, New Zealand. This partici- pant pool was recruited via birth registrations of infants born in the Dunedin area. The age groups were 5-year-old children (M � 5.42 years, SD � 0.32 years), 8- to 9-year-old children (M � 8.99 years, SD � 0.40 years), 12- to 13-year-old adolescents (M � 12.96 years, SD � 0.37 years). The adults were first- and second-year psychology undergraduate students (18- to 20-year-olds) at the University of Otago (M � 19.17 years, SD � 0.88 years).1 There were six males and six females in each age group. Consent for the children to participate was obtained from their parents, and inde- pendent consent was obtained from the adolescents and the adults.
Materials
For each participant, a Timeline template was made in advance using colored card (red, purple, or turquoise; 101.8 cm � 66.2 cm). A line measuring 80 cm was drawn horizontally in the center of the card. Vertical lines marking each year of the participant’s life were drawn at equal intervals along the horizontal line and measured 4 cm in length. The corresponding age was written beneath each vertical line (e.g., “[participant’s name] is born!” “1 year old,” “2 years old,”“[participant’s present age] years old. Today!”).
The participants’ parents were asked to provide photos of the participant at various ages. For the 5-year-olds, parents were asked to provide a photo of the participant as a newborn, as a 1-year-old, and as a 3-year-old. For the 8- to 9-year-olds, parents were asked to provide a photo of the participant as a newborn, as a 1-year-old, as a 3-year-old, and as a 5-year-old. For the 12- to 13-year-olds, parents were asked to provide a photo of the participant as a newborn, as a 1-year-old, as a 3-year-old, as a 5-year-old, and as a 10-year-old. Finally, for the 18- to 20-year-olds, parents were asked to provide a photo of the participant as a newborn, as a 1-year-old, as a 3-year-old, as a 5-year-old, as a 10-year-old, as a 13-year-old, and as a 15-year-old. In addition, for all age groups, the experimenter took a photo of the participant using a digital camera on the day of the interview. None of the photos that were used in the Timeline included additional cues (e.g., objects/ locations), nor did the photos represent the memories that were
1 Information about the languages spoken at home and socioeconomic status was not collected for this sample; however, the socioeconomic profile of this area of New Zealand is relatively flat. All of the participants would have been classified as middle to upper middle class.
1051AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
subsequently discussed during the interview. When invited to participate, the parents of participants were told that our purpose in the study was to understand age-related changes in memory. They were not told anything about our interest in the participants’ earliest memories. Parents were specifically asked not to show their children the photos in advance of the interview. To the best of our knowledge, all parents adhered to our requests.
Procedure
All participants were tested in a quiet room in the Early Learn- ing Project Laboratory. Parents accompanied their children to the laboratory. Once the child felt comfortable, the parent went into another room with a second experimenter. The adult participants attended the interview alone.
At the beginning of the interview, the experimenter took the participant’s photo using a digital camera. For the three younger groups of participants, the experimenter gave the new photo to the participant with the photos that his or her parent had provided and asked the participant to place the photos in order from youngest to oldest. Once the participant had performed this task, the experi- menter presented him or her with the Timeline and asked the participant to glue his or her photos onto the Timeline at the appropriate ages. A line extending from each target age was drawn onto the Timeline in advance, to make it clear where the photos should be located. For the adult participants, the procedure differed only insofar as their photos were attached to their Timelines by the experimenter in advance of the interview. Sample Timelines for each age group are shown in Figure 1.
The experimenter began each interview by explaining to the participant that he or she would be asked to describe some mem- ories from the past. The participant was instructed to tell the experimenter only the details of his or her memories that he or she
actually remembered and not to fill in any blanks with details that he or she had obtained from family stories or photos and videos.
Once the interview procedure had been explained, the experi- menter reviewed the sequence of the Timeline with the participant. For example, the experimenter stated that the Timeline started at the participant’s birth and stopped at the participant’s present age. The experimenter then pointed to the photos at each of the target ages, saying, “Look, this is a photo from when you were born” to “And this is you today, [present age] years old.” The Timeline activity was performed to reinforce the notion of a linear time sequence, particularly for the younger children, and to provide an external cue for thinking about memories from different epochs of the participant’s life.
Next, the experimenter asked the participant to think of some- thing that he or she remembered happening or doing during the past month. The experimenter pointed to the photo of the partic- ipant at the present age. Once the participant had nominated an event, the experimenter first gave him or her opportunity for free recall by providing general prompts (e.g., “Tell me about [the event],” followed by “Can you remember anything else about [the event]”) until the participant indicated that he or she did not have any more information to report. The experimenter then asked a set of four specific questions (directed recall) about the memory. The questions were “Who else was there?” “Where were you?” “What did you do?” and “How did you feel?” The questions were followed by another opportunity for free recall (e.g., “Is there anything else that you can remember now that you would like to tell me about [the event]?”), until the participant could recall no more.
After this recent memory had been recalled, the experimenter asked the participant to report one event from when he or she was 3 years old, one event from before he or she was 3 years old, and
Figure 1. Sample Timelines for (clockwise from top left) 5-year-olds, 8- to 9-year-olds, 18- to 20-year-olds, and 12- to 13-year-olds. These photos have been published with the written permission of the participants and their parents.
1052 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
his or her earliest memory. The earliest memory was explained as “Your first memory” or “The first thing that you remember doing or happening to you.” The recollection of all of the events took the same format as did recollection of the recent event (i.e., free recall followed by specific questions and ending with free recall). The participant was asked to date his or her earliest memory and his or her memory from before the age of 3 years by placing a sticker on the Timeline to indicate the age at which the event took place. The experimenter also asked the participant to verbally clarify the age. Participants (including the children) typically answered this clar- ification question using an age and a fraction (e.g., 2 1⁄4, 3 1⁄2). This age was then converted into months (e.g., 27 months, 42 months, respectively) for parental verification and for statistical analysis.
While the participant completed this portion of the interview, a second experimenter asked his or her parent to nominate two salient events in the child’s life for each of the target ages (i.e., recent, 3 years old, before 3 years old). The experimenter briefly transcribed these events (without specific detail) for the experi- menter interviewing the participant. Because parents did not ac- company the 18- to 20-year-olds to their interviews, an experi- menter visited their parents in advance of the interview to obtain these parent-nominated events.
Once the participant had recalled his or her earliest memory, the experimenter asked whether he or she remembered one parent- nominated event per target age, from most to least recent. Where the participant and parent had both nominated the same event, the exper- imenter asked about an event different from the self-nominated event. If the participant stated that he or she did not remember an event that his or her parent had nominated, the second alternative was used (where it was different from what the participant had previously reported). Again, the same interview structure was used for the parent-nominated events as for the self-nominated and earliest events. The entire interview was video- and audiotaped.
The interviews were transcribed verbatim to determine the content of participants’ memories. The transcript of the interview was then presented to one of the participant’s parents. Parents evaluated the content of the transcripts within 2–6 months of the original interview. Parents were asked to first read over the memory and decide if the event had definitely happened, may have happened, or had definitely never happened. If the event had definitely happened, the parents were asked to indicate whether they had some firsthand knowledge of it. If the parents did have some firsthand knowledge, they were asked to consider each piece of information (i.e., each line of the transcript) and rate it as correct, possible, or incorrect. In this way, only parents who knew that the event had definitely happened and had some firsthand knowledge of it rated the accuracy of the details provided by their child during the interview. The parents were also asked to indicate whether the estimation by their child of his or her age at the time of the event was correct or incorrect. If the age was incorrect, they were asked to note the correct age. As mentioned previously, the participants dated their earliest memories by placing a sticker on the Timeline at the age that the event took place. In general, participants dated their memories within blocks of 3 months (i.e., by quarter years).
Coding
To assess potential age-related differences in the amount of information reported, we parsed all of the participant’s memories
into individual clauses. A clause corresponded roughly to a simple sentence and was defined by the presence of a verb, such that there was one verb per clause (see Gross & Hayne, 1998). Any infor- mation that was unrelated to the description of the event was not included. Similarly, repetitions of the same information were not included (e.g., where the participant provided the same informa- tion in response to the specific questions that he or she had provided during free recall). The numbers of clauses provided by the participant during free recall and during directed recall were summed to produce a score for the total amount of information provided. A second coder independently parsed 25% of each group’s memories into clauses (n � 64 memories). Pearson product–moment correlations yielded interrater reliability coeffi- cients of .90 for total recall, .91 for free recall, and .85 for directed recall.
Recall that the defining feature of episodic memory is auto- noetic awareness, that is, the rememberer reexperiences the event as having happened to “me” (Tulving, 2002a, 2002b). To estimate the autonoetic content of the information reported, we devised a coding scheme in which the individual clauses of information that were reported during free recall were assigned to one of five mutually exclusive categories. The categories were (a) autonoetic content—clauses in which the participant referred to himself or herself by using the first person (i.e., “I” or “we” statements); (b) second-person references—clauses in which the participant re- ferred to him or herself by using the second person (i.e., “you” statements); (c) past tense—clauses in which the participant used an explicit verb in the past tense; (d) present tense—clauses in which the participant used an explicit verb in the present tense; and (e) lists—clauses in which the participant did not use an explicit verb but listed pieces of information. The categories were exhaus- tive. There were instances in which specific clauses could techni- cally be assigned to more than one category (e.g., “I pushed the button” could be coded both as autonoetic content and as past tense). In these instances, the clauses were always coded in the following order: (1) autonoetic content, (2) second-person refer- ences, (3) past tense, (4) present tense, and (5) lists. In addition, if a sentence was split into two or more clauses and the first clause contained an autonoetic component or a second-person reference, the subsequent clause was always coded the same as the first. For example, “We went to the circus and saw some elephants” would contain two autonoetic clauses: “We went to the circus” and “and saw some elephants.” A second “we” is implied in the second clause. A second coder independently assigned the clauses from 25% of each group’s memories to the five categories (n � 593 clauses). Agreement between the coders was 87.18% (Cohen’s � � .79). Note that of the five categories, autonoetic content and past tense reflect typical episodic memory and second-person references, present tense, and lists reflect typical semantic mem- ory.
Results
Recall that participants were asked to describe one self- nominated and one parent-nominated memory for each of the three target ages (recent, 3 years old, before 3) in addition to their earliest memory. With the exception of memories for an event that occurred recently and participants’ earliest memories, memories that the participants’ parents indicated had occurred after the age
1053AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
of 3 years were excluded from the subsequent analyses (n � 8: 5-year-olds � 1; 8- to 9-year-olds � 5; 12- to 13-year-olds � 1; 18- to 20-year-olds � 1). One additional memory was excluded because the 5-year-old participant’s parent indicated that the event did not take place. Two recent memories provided by different 5-year-old children were also excluded because they were of generic, repeated, everyday events. The 5-year-old children also provided an additional 5 memories that had occurred between the ages of 3 and 5 years. These memories were provided spontane- ously and were excluded from the final sample.
There were 231 memories in the final sample. These memories were assigned to one of four memory categories based on the age at which the event occurred: (a) Recent memories included mem- ories that occurred within the past month; (b) 3 years included memories that occurred between ages 3 and 4 years; (c) under 3 years included memories that occurred under the age of 3 years; and (d) earliest contained participants’ earliest memory. In theory, each participant could report a maximum of 7 memories (2 recent, 2 from age 3, 2 from under age 3, and 1 earliest). Because not all participants could remember events from each of these categories (with the exception of the recent memories category), the number of memories provided by each participant varied. Note, however, that the absolute number of memories that were reported were equally distributed across age groups for both recent memories (5-year-olds � 22; 8- to 9-year-olds � 24; 12- to 13-year-olds � 24; 18- to 20-year-olds � 24) and for memories at the age of 3 or below and earliest memories (5-year-olds � 35; 8- to 9-year- olds � 36; 12- to 13-year-olds � 33; 18- to 20-year-olds � 33), indicating that there were no age-related differences in the ability of participants to generate their own memories or to recall the memories that were provided by their parents.
Dating Consistency
Parents were able to verify the ages of 221 of the 231 memories in the final sample (95.57%). First, in order to assess potential age-, gender-, and memory category-related differences in whether or not the participants dated their memories correctly, we con- ducted a series of chi-square tests on the number of memories that participants dated correctly as a function of age group, gender, and memory category (recent memory, 3 years, under 3 years, earliest). Overall, participants were consistent with their parents in estimat- ing their age at the time of the event; participants and parents agreed on 87.33% (n � 193) of the memories. There were no differences with respect to gender, �2(1, N � 221) � .01, p � .05, V � 0.01, 95% CI [0.07, 0.14], power � .99, or age group, �2(3, N � 221) � 3.04, p � .05, V � 0.10, 95% CI [0.10, 0.24], power � .97.2 In contrast, there were differences in the number of consistently dated memories as a function of memory category, �2(3, N � 221) � 45.14, p � .01, V � 0.45, 95% CI [0.33, 0.58]; perhaps not surprisingly, participants and parents had more dis- agreements when they were asked to date the earliest memories (67.50% correct) and memories from under 3 years (64.71% correct) than they did when dating memories from age 3 (96.49% correct) and recent memories (98.89% correct), smallest �2(1, N � 97) � 15.11, p � .01, V � 0.40, 95% CI [0.22, 0.60].
Although participants and parents were consistent when asked to date the memories, they did have some disagreements. In order to determine the magnitude of these dating differences, we sub-
tracted the parents’ estimates of age (in months) from the partic- ipants’ estimates of age (in months) for the inconsistently-dated memories only (n � 28). Overall, participants’ mean difference of estimation was only 0.63 months (SD � 13.81 months). Although participants slightly underestimated the ages at which the events occurred, their mean difference of estimation was not significantly different from zero (perfectly consistent dating), t(27) � 0.24, p � .05, d � 0.05, 95% CI [�0.33, 0.42], power � .84.
Parents’ corrected age estimates were taken as the age of the memory in all subsequent analyses, in order to provide the most conservative estimate of the age of participants’ memories. The participant’s estimate of the age of the memory was used in cases in which parents did not have firsthand knowledge of the memory (n � 10).
Accuracy of Details
Parents had at least some firsthand knowledge of 174 of the 231 memories in the final sample (75.32%). As judged by parents, the content of the participants’ reports of these memories was highly accurate. On average, parents rated 76.56% (SD � 23.08%) of the details as correct, 18.48% (SD � 21.46%) of the details as possi- ble, and only 5.60% (SD � 12.64%) of the details as incorrect. The most conservative measure of accuracy was used in subjecting the proportions of details that parents rated as correct to a 2 (Gen- der) � 4 (Age Group: 5-year-olds, 8- to 9-year-olds, 12- to 13-year-olds, 18- to 20-year-olds) � 4 (Memory Category: recent memory, 3 years, under 3 years, earliest memory) analysis of variance (ANOVA). The proportion of details that parents rated as correct did not vary as a function of gender, F(1, 142) � 0.18, p � .05, 2 � .001, 95% CI [0, .04], power � .85; age group, F(3, 142) � 2.17, p � .05, 2 � .02, 95% CI [0, .07], power � .71; or memory category, F(3, 142) � 0.83, p � .05, 2 � .01, 95% CI [0, .04], power � .71. There were no interactions, largest F(3, 142) � 1.96, p � .05, 2 � .04, 95% CI [0, .10], power � .71.
Distribution of Early Memories
Next, we assigned the participants’ earliest memory and mem- ories from under the age of 3 years (n � 78) to one of four categories (0–1 years, 1–2 years, 2–3 years, 3� years), depending on the age at which the event occurred. Table 2 shows the pro- portion of these early memories that fell into each age band as a function of the participant’s age at the time of the interview. In order to assess potential age-related differences in the distribution of these early memories, we conducted a chi-square test on the number of memories that fell into each age band (0–1 years, 1–2 years, 2–3 years, 3� years) as a function of age group. The number of memories that fell into each age band differed as a function of participant age, �2(9, N � 78) � 25.68, p � .01, V � 0.33, 95%
2 For all chi-square tests, t tests, and ANOVAs, Cramér’s V (Cramér, 1999), Cohen’s d (Cohen, 1988), and eta-squared (Cohen, 1973) measures of effect size (with corresponding 95% confidence intervals), respectively, have been included. In addition, the power of the test at detecting a medium effect (Cohen’s w � .30, d � 0.50, f � 0.25) is provided for all nonsig- nificant effects (Cohen, 1988). These power values represent the probabil- ity that a statistical test will detect a statistically significant effect should a medium-sized or larger effect actually exist.
1054 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
CI [0.22, 0.46]. In order to clarify this effect, we conducted a series of chi-square tests on the number of memories reported by each age group separately for each age band. The number of memories that fell into the 2–3 years age band and the 3� years age band did not differ as a function of participant age, largest �2(3, N � 15) � 6.07, p � .05, V � 0.63, 95% CI [0.45, 1.15], power � .33. There were, however, effects of participant age on the number of mem- ories that fell into the 0–1 years age band, �2(3, N � 8) � 9.00, p � .05, V � 1.06, 95% CI [0.62, 1.77] and the 1–2 years age band, �2(3, N � 14) � 8.29, p � .05, V � 0.77, 95% CI [0.46, 1.30]. Within the 0–1 years age band, the 8- to 9-year-olds provided a greater proportion of memories (23%) than did the 12- to 13-year-olds (0%) or the 18- to 20- year-olds (0%), smallest �2(1, N � 5) � 5.00, p � .05, V � 1.00, 95% CI [0.45, 1.93]. The 5-year-olds also provided a greater proportion of memories in this band (21%) than did the 12- to 13-year-olds and the 18- to 20-year-olds, but these differences were not significant, largest �2(1, N � 3) � 3.00, p � .05, V � 1.00, 95% CI [0.57, 2.21]. Within the 1–2 years age band, the 12- to 13-year-olds provided a greater proportion of memories (42%) than did the 18- to 20-year- olds (4%), �2(1, N � 9) � 5.44, p � .05, V � 0.77, 95% CI [0.35,
1.47]. The 5-year-olds and the 8- to 9-year-olds provided the same proportion of memories in this band (21% and 9%, respectively) as did the 12- to 13-year-olds and the 18- to 20-year-olds, largest �2(1, N � 10) � 3.60, p � .05, V � 0.60, 95% CI [0.32, 1.26].
Note that, on the whole, the vast majority of the 5-year-olds’ and the 8- to 9-year-olds’ early memories fell before the age of 3 years (92% and 91%, respectively). In contrast, 74% of the 12- to 13-year-olds’ early memories and 70% of the 18- to 20-year-olds’ early memories fell before the age of 3 years. Note also that only the 5-year-olds and the 8- to 9-year-olds reported early memories that fell into the 0–1 years age band. All of the groups reported some memories that fell into the 1–2 years age band, but this number was very small for the 18- to 20-year-olds (4%); in fact, only one adult was able to report one memory in this age band. This finding is highly consistent with prior research conducted with adults. Overall, approximately 3% of adults’ early memories (from before the age of 5) fall before the age of 2 years (see Rubin, 2000). Thus, although all of the groups reported early memories, the children reported a greater proportion of younger early mem- ories than did the adolescents, and the adolescents, in turn, re- ported a greater proportion of younger early memories than did the adults.
Average Age of Earliest Memories
In addition, we were interested in the participants’ very earliest personal memory. The average age of participants’ earliest mem- ories is shown in Figure 2 as a function of age group. The age of participants’ earliest memory was subjected to a 2 (Gender) � 4 (Age Group) ANOVA. There was a main effect of age group, F(3, 33) � 8.24, p � .01, 2 � .43, 95% CI [0.13, 0.58]. As shown in Figure 2, the 5-year-old children, the 8- to 9-year-old children, and
Figure 2. The average age of participants’ earliest memories (�1 SE) as a function of age group as compared to the 3 1⁄2-year boundary of childhood amnesia (as indicated by the solid line at 3.5 years). The asterisks indicate that the average age of the participants’ earliest memory was below the 3 1⁄2-year boundary of childhood amnesia ( p � .05).
Table 2 The Proportion of Early Memories Occurring at 0–1 Years, 1–2 Years, 2–3 Years, and 3� Years as a Function of Age Group
Age group 0–1 1–2 2–3 3�
5-year-olds 0.21 0.21 0.50 0.07 8- to 9-year-olds 0.23 0.09 0.59 0.09 12- to 13-year-olds 0 0.42 0.32 0.26 18- to 20-year-olds 0 0.04 0.65 0.30
1055AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
the 12- to 13-year-old adolescents all had younger earliest mem- ories than did the 18- to 20-year-old adults, smallest t(22) � 2.17, p � .05, ds � 1.12, 1.78, and .89, respectively, 95% CIs [0.03, 2.16], [0.72, 2.80], and [0.04, 1.72], respectively. There was also a main effect of gender, F(1, 33) � 11.09, p � .01, 2 � .25, 95% CI [0.04, 0.46]; overall, the male participants reported younger earliest memories (M � 1.71 years, SD � 1.42 years) than did the female participants (M � 2.70 years, SD � 1.16 years). There was no interaction, F(3, 33) � 2.78, p � .05, 2 � .20, 95% CI [0, .37], power � .20.
Recall that when adults are asked to nominate their earliest personal memory, the average age at which that event occurred is 3 1⁄2 years (see Table 1). In order to determine whether the Timeline procedure altered our estimates of the age of earliest memory in adults, we conducted single-sample t-tests between the age of earliest memories and the 3 1⁄2-year boundary of childhood amnesia that is typically reported in the literature. These analyses indicated that the average age of earliest memory for the 18- to 20-year-olds did not differ from the typical 3 1⁄2-year boundary, t(11) � 1.05, p � .05, d � 0.30, 95% CI [�0.28, 0.88], power � .52. In contrast, the average age of earliest memory for the 5-year- olds, the 8- to 9-year-olds, and the 12- to 13-year-olds was signif- icantly younger than the traditional 3 1⁄2-year boundary of child- hood amnesia in adults, smallest t(6) � 3.06, p � .05, ds � 1.16, 1.82, and 1.18, respectively, 95% CIs [0.15, 2.11], [0.77, 2.83], and [0.42, 1.92], respectively.
Amount of Information
Overall, the Timeline procedure elicited rich accounts of par- ticipants’ early memories. To assess potential differences in the amount of information that participants reported, we subjected the total number of clauses that participants reported during free recall to a 2 (Gender) � 4 (Age Group) � 4 (Memory Category) ANOVA. As shown in Table 3, participants reported an average of 12.64 clauses (SD � 13.50 clauses) during free recall, and this value did not differ as a function of gender, F(1, 199) � 0.21, p � .05, 2 � .001, 95% CI [0, 0.03], power � .94. There was, however, a significant effect of age group, F(3, 199) � 2.96, p � .05, 2 � .04, 95% CI [0, 0.10]; the 18- to 20-year-olds provided more information than did the 5-year-olds and the 8- to 9-year- olds, smallest t(101) � 3.62, p � .05, ds � 0.97 and 0.67, respectively, 95% CIs [0.57, 1.36] and [0.29, 1.04], respectively. The 12- to 13-year-olds also provided more information than did the 5-year-olds, t(64) � 2.08, p � .01, d � 0.39, 95% CI [0.01, 0.76].
There was also a main effect of memory category, F(3, 199) � 6.12, p � .01, 2 � .08, 95% CI [0.02, 0.15]; participants provided almost twice as much information about recent memories as they did about their earliest memories, t(133) � 2.74, p � .01, d � 0.51, 95% CI [0.14, 0.88], memories that occurred under the age of 3 years, t(125) � 3.69, p � .01, d � 0.72, 95% CI [0.32, 1.11], and memories that occurred at age 3, t(122) � 3.66, p � .01, d � 0.61, 95% CI [0.27, 0.94]. There were no significant interactions, largest F(3, 199) � 2.50, p � .05, 2 � .04, 95% CI [0, 0.09], power � .86.
Episodic Content
Recall that one way in which we evaluated the episodic content of the memories was to assign the clause data from the free recall portion of the interview to one of five categories (autonoetic content, second-person references, past tense, present tense, and lists). We combined the values for autonoetic content and past tense to reflect episodic content, and we combined the values for second-person references, present tense, and lists to reflect seman- tic content. These combinations are consistent with Tulving’s (1972, 2002a, 2002b) definition of episodic and semantic memory, respectively. Because the absolute amount of information that participants provided during free recall differed as a function of participant age and when the event occurred (i.e., recent, 3 years, under 3 years, earliest), we converted the raw values into propor- tions for each coding category.
Overall, the majority of clauses were coded as episodic (M � 85.56%, SD � 17.10%; autonoetic: M � 51.74%, SD � 24.89%; past tense: M � 33.81%, SD � 23.54%) rather than semantic (M � 14.48%, SD � 17.13%; second-person references: M � 2.00%, SD � 8.25%; present tense: M � 6.80%, SD � 11.10%; lists: M � 5.69%, SD � 11.08%). The proportion of episodic content was subjected to a 2 (Gender) � 4 (Age Group) � 4 (Memory Cate- gory) ANOVA. As shown in Table 4, the proportion of episodic content did not differ as a function of participant age, F(3, 199) � 0.69, p � .05, 2 � .01, 95% CI [0, 0.04], power � .86, or the age of the memory (i.e., earliest, under 3, 3 years, recent), F(3, 199) � 0.36, p � .05, 2 � .005, 95% CI [0, .03], power � .86. There was no effect of gender, F(1, 199) � 0.48, p � .05, 2 � .002, 95% CI [0, .03], power � .98, and no interactions, largest F(3, 199) � 1.63, p � .05, 2 � .02, 95% CI [0, .07], power � .86.
Recall that the other way in which we evaluated the episodic content of the memories was to examine the ability of participants to answer the specific questions regarding their memories (i.e., “who,” “what,” “where”). In order to do this, we compared the number of participants in each age group who provided at least one
Table 3 Average Number of Free-Recall Clauses for Each Age Group as a Function of the Age of the Memory
Age group Earliest Under 3 3 years Recent Total
5-year-olds 7.57 (3.99) 7.57 (5.53) 8.29 (6.12) 9.09 (6.68) 8.42 (5.96)a,c
8- to 9-year-olds 7.20 (6.03) 8.75 (7.56) 8.50 (5.84) 14.08 (8.83) 10.57 (7.92)b
12- to 13-year-olds 8.33 (5.29) 7.71 (2.29) 11.21 (7.27) 21.88 (32.03) 14.67 (21.90)c
18- to 20-year-olds 17.42 (11.07) 11.18 (5.42) 10.20 (5.05) 22.33 (12.18) 17.02 (11.02)a,b
Total 10.59 (8.48)d 9.05 (5.81)e 9.36 (6.16)f 17.01 (18.72)d,e,f 12.64 (13.50)
Note. Standard deviations are in parentheses. Superscripts of the same letters indicate significant differences ( p � .05) between the respective values.
1056 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
piece of information in each question category (during either free recall or directed recall). Overall, participants provided informa- tion regarding all three of the specific questions for 96.54% of all of the memories. Furthermore, there was no difference in the proportion of questions that participants answered as a function of question category (i.e., who � 98.27%, where � 97.84%, what � 100%), �2(2, N � 693) � 4.73, p � .05, V � 0.10, 95% CI [0.05, 0.17], power � .99.
Next, we assigned each memory a score (0–3) based on the number of questions that the participant was able to answer about that memory. Table 4 shows the average number of questions that participants answered as a function of participant age and memory age. The number of questions that participants answered was subjected to a 2 (Gender) � 4 (Age Group) � 4 (Memory Cate- gory) ANOVA. As shown in Table 4, participants answered, on average, 2.96 questions (SD � 0.22 questions), and this figure did not differ as a function of the age of the memory (i.e., earliest, under 3, 3 years, recent), F(3, 199) � 0.78, p � .05, 2 � .01, 95% CI [0, .04], power � .86, or as a function of gender, F(1, 199) � 1.34, p � .05, 2 � .01, 95% CI [0, .05], power � .94. There was, however, an effect of participant age, F(3, 199) � 3.52, p � .05, 2 � .05, 95% CI [.002, .11]; as shown in Table 4, the 8- to 9-year-olds answered fewer questions than did the 18- to 20-year- olds, t(59) � 2.19, p � .05, d � 0.41, 95% CI [.03, .77]. Note, however, that participants of all ages were able to answer the majority of the questions about their memories (i.e., more than 2) regardless of when the events occurred. There were no interac- tions, largest F(9, 199) � 1.37, p � .05, 2 � .06, 95% CI [0, .09], power � .69.
Other Sources of Information
In nominating each of their memories, participants were asked to describe events that they recalled rather than events that they knew about exclusively on the basis of photos or family stories. This instruction did not preclude the possibility that, in addition to firsthand recollection, there might be photos or stories about the target events available. In order to assess the potential effects of these other sources of information, we asked participants and parents whether there were photos or family stories about each
participant’s memories at the conclusion of the interview. We used these data to conduct a number of additional analyses.
First, in order to assess potential age-related differences in the presence of photos or family stories, we conducted a series of chi-square tests on the number of memories for which the partic- ipants had photos or family stories available both as a function of age group and as a function of memory category. These analyses yielded no differences with respect to memory category, �2(3, N � 223) � 3.16, p � .05, V � 0.10, 95% CI [0.10, 0.24], power � .97. Irrespective of the period during which the memory occurred (i.e., recent, 3 years, under 3 years, earliest), there were photos or family stories available for 73.54% of the memories. In contrast, there were age-related differences in the number of photos or family stories available about the memories, �2(3, N � 223) � 17.77, p � .01, V � 0.28, 95% CI [0.17, 0.41]; the 5-year-olds had a greater proportion of photos or stories available about their memories (91.23%) than did the 8- to 9-year-olds (60.38%) or the 12- to 13-year-olds (63.16%), smallest �2(1, N � 114) � 12.76, p � .01, Vs � 0.36 and 0.33, respectively, 95% CIs [0.20, 0.56] and [0.17, �0.53], respectively. The 18- to 20-year-olds also had a greater proportion of photos or stories available (78.57%) than did the 8- to 9-year-olds, �2(1, N � 109) � 4.27, p � .05, V � 0.20, 95% CI [0.10, 0.40].
Next, in order to evaluate whether the presence of photos or family stories made a difference to the age, amount of information, episodic content (number of “who,” “where,” and “what” ques- tions answered and proportion of episodic content), and accuracy of participants’ memories, we conducted a series of 4 (Age Group) � 4 (Memory Category) � 2 (Reminder: Photos or Stories, No Photos or Stories) ANOVAs. The results of these analyses indicate that there were no effects (or interactions) of photos or family stories on the age of the memories, largest F(3, 192) � 2.22, p � .05, 2 � .03, 95% CI [0, .08], power � .84; the amount of information reported during free recall, largest F(1, 192) � 1.00, p � .05, 2 � .01, 95% CI [0, .04], power � . 94; the number of questions answered, largest F(3, 192) � 1.09, p � .05, 2 � .02, 95% CI [0, .05], power � .84; the proportion of episodic content, largest F(1, 192) � 3.18, p � .05, 2 � .02, 95% CI [0,
Table 4 Average Proportion of Episodic Clauses and Average Number of Questions Answered for Each Age Group as a Function of the Age of the Memory
Age group Earliest Under 3 3 years Recent Total
Episodic content 5-year-olds 0.78 (0.21) 1.00 (0) 0.84 (0.18) 0.81 (0.25) 0.84 (0.21) 8- to 9-year-olds 0.91 (0.13) 0.85 (0.19) 0.83 (0.23) 0.87 (0.16) 0.86 (0.18) 12- to 13-year-olds 0.83 (0.21) 0.86 (0.16) 0.84 (0.13) 0.83 (0.15) 0.84 (0.16) 18- to 20-year-olds 0.87 (0.14) 0.84 (0.13) 0.96 (0.06) 0.87 (0.14) 0.88 (0.13) Total 0.85 (0.17) 0.88 (0.15) 0.86 (0.17) 0.85 (0.18) 0.86 (0.17)
No. questions answered (max � 3) 5-year-olds 3.00 (0) 3.00 (0) 2.95 (0.22) 2.95 (0.21) 2.96 (0.19) 8- to 9-year-olds 2.90 (0.32) 2.67 (0.65) 2.93 (0.27) 3.00 (0) 2.90 (0.35)a
12- to 13-year-olds 3.00 (0) 3.00 (0) 2.93 (0.27) 3.00 (0) 2.98 (0.13) 18- to 20-year-olds 3.00 (0) 3.00 (0) 3.00 (0) 3.00 (0) 3.00 (0)a
Total 2.98 (0.16) 2.89 (0.39) 2.95 (0.22) 2.99 (0.10) 2.96 (0.22)
Note. Standard deviations are in parentheses. Superscripts of the same letters indicate significant differences ( p � .05) between the respective values.
1057AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
.07], power � .94; or the accuracy of the memories, largest F(1, 139) � 1.49, p � .05, 2 � .03, 95% CI [0, .09], power � .69.
Discussion
In the present experiment, we used our newly developed Time- line procedure to compare the boundary of childhood amnesia in children, adolescents, and adults. We found that although partici- pants of all ages reported some very early memories (under 3 years), the children reported a greater proportion of younger early memories than did the adolescents. The adolescents, in turn, re- ported a greater proportion of younger early memories than did the adults. When we examined the single earliest memory that was reported by the participants, we found that the children and ado- lescents reported younger earliest memories than did adults (see also Cleveland & Reese, 2008; Jack, MacDonald, Reese, & Hayne, 2009). Furthermore, although the average age of the adults’ single earliest memories did not differ from the 3 1⁄2-year boundary typically reported in the literature, the average age of the children’s and adolescents’ earliest memories was significantly younger than 3 1⁄2 years.
When we examined the content of the memories, we found that participants of all ages were able to provide specific information about the “who,” “what,” “where,” and “when” of their memories regardless of the age at which those memories occurred. In addi- tion, the vast majority of the information that participants provided was episodic rather than semantic and the proportion of episodic content that was reported was the same irrespective of the age of the participant or the age of the memory. Although many partici- pants had some form of reminder (photos or family stories) about some of the memories that they reported, the presence of a re- minder did not affect the age, accuracy, amount of information, or episodic content of these memories.
The current data shed new light on the phenomenon of child- hood amnesia. Although this phenomenon has been recognized for more than a century, the source of this amnesia is not well understood. Freud’s (1905/1963) original explanation involved repression. He argued that the memories of our early experiences were stored in pristine condition but were blocked from conscious- ness (i.e., repressed), due to their aggressive and sexual content. Given the overwhelming lack of empirical evidence for the phe- nomenon of repression (Hayne, Garry, & Loftus, 2006), research- ers have turned to other potential explanations of childhood am- nesia (for a review, see K. Nelson & Fivush, 2004). These include, but are not limited to, incomplete neurological maturation (Bachevalier, 1992; Nadel & Zola-Morgan, 1984; C. A. Nelson, 1998), immature cognitive ability (Pillemer & White, 1989), an immature sense of self (Howe & Courage, 1993, 1997; Welch- Ross, 2000), parent–child conversations about the past (Fivush & Nelson, 2004; K. Nelson, 2003; K. Nelson & Fivush, 2004), language development (Cheatham & Bauer, 2005; Howe, Courage, & Peterson, 1994; K. Nelson, 1996; K. Nelson & Fivush, 2004; Richardson & Hayne, 2007; Simcock & Hayne, 2002, 2003), and cognitive and social discontinuity between early childhood and adulthood (Neisser, 1962; Schachtel, 1947; Wang, 2001). The present data bear on two additional explanations of childhood amnesia that have received comparatively little experimental at- tention.
Episodic Memory and Childhood Amnesia
Tulving (1972, 2002a, 2002b) and others (Perner & Ruffman, 1995; Suddendorf & Corballis, 1997) have considered childhood amnesia within the context of other forms of human amnesia. These researchers have argued that infants and young children lack episodic memory skill and that this precludes them from forming autobiographical memories in the first place. According to this perspective, children under the age of 4 years do not possess the late-developing episodic memory system that is required to encode the who, what, where, and when of their experiences. More re- cently, these theorists have also argued that young children are incapable of autonoetic consciousness, or chronesthesia, the feel- ing that one has experienced the specific event in the past. This type of autonoetic consciousness is sometimes described as mental time travel (Tulving, 2002a, 2002b) and is a fundamental charac- teristic of autobiographical memory in adults.
The present findings challenge the view that episodic memory does not initially emerge until after the age of 4. For example, the children and the adolescents in the present study accurately re- ported information about the who, what, when, and where of events that took place even when they were much younger than 4 years old. The vast majority of the information that these children and adolescents provided about their memories was consistent with Tulving’s definition of episodic memory; the proportion of episodic content was the same irrespective of the age of the participant or the age of the memory. Thus, if we accept adult recollection as the benchmark for episodic memory (Tulving, 1972, 2002a, 2002b), and if children’s recollections resemble adults’ recollections, we must conclude that young children can encode, retain, and recall their past experiences in an episodic manner. This conclusion is highly consistent with a large body of literature showing that 2- and 3-year-old children report episodic information if they are interviewed during early childhood per se (Fivush et al., 1987; Fivush & Hamond, 1990; Fivush, Hamond, Harsch, Singer, & Wolf, 1991; MacDonald & Hayne, 1996; K. Nelson, 1992; Peterson, 2002; Peterson & Rideout, 1998). Taken together, the present study and prior prospective studies of mem- ory development in young children suggest that although episodic memory ability clearly increases over the course of development, childhood amnesia cannot be explained by a lack of episodic memory during early childhood.
Forgetting and Childhood Amnesia
Although many theorists have traditionally discounted forget- ting as an important mechanism in childhood amnesia (e.g., Cro- vitz & Harvey, 1979; Crovitz et al., 1980; Crovitz & Quina- Holland, 1976; Rubin, 1982, 2000; Waldfogel, 1948), the present findings suggest that forgetting may play at least some role in our inability to recall our early experiences (Jack et al., 2009; Morrison & Conway, 2010; Peterson et al., 2005; Wetzler & Sweeney, 1986). On the basis of both prospective and retrospective data, we argue that young children begin to encode episodic memories very early in development and at least some of these memories remain accessible until early adolescence. Over time, however, access to these memories is gradually lost because the information is too lean or too poorly organized to be retrieved over very long delays (Brainerd & Reyna, 1990; Morrison & Conway, 2010; Reyna & Brainerd, 1995).
1058 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Data obtained from other studies conducted with adult samples provide additional support for the role of forgetting in childhood amnesia. For example, when adults are repeatedly probed with an exhaustive interview technique to recall earlier and earlier mem- ories (see Crovitz & Harvey, 1979; Waldfogel, 1948), they tend to report younger earliest memories than when they are asked to provide a single earliest memory (Wang et al., 2004; Wright, 2006). This finding suggests that some early memories may remain available in storage but that they are less accessible than more detailed memories that were encoded later in development. With additional cues and prompts, however, adults may regain access to these very young and very lean representations.
Up to this point we have assumed that the present pattern of results reflects age-related changes in the boundary of childhood amnesia, but a number of alternative explanations should also be considered. First, it is possible that there were dating errors in the age of participants’ earliest memories. This problem is inherent in all retrospective research of this kind, even when that research is conducted exclusively with adults. In the present experiment, we first asked participants to date their memories and then asked parents to verify the accuracy of those dates. Although there were no age-related differences in the consistency of participants’ dates, in the end, we used parents’ estimates in all of our analyses. Thus, in order for dating errors to influence our conclusions regarding age-related differences in childhood amnesia, there would have to be systematic dating errors in parents’ accuracy as a function of the age of the child. In our view, this is highly unlikely.
Second, the age-related differences in childhood amnesia re- ported here may also reflect age-related differences in source monitoring (Johnson, Hashtroudi, & Lindsay, 1993; Johnson & Raye, 1981). Although all participants were explicitly told to report only information that they remembered on the basis of their prior experience, it is still possible that they reported information that they had gleaned from family stories and photographs. Given that children’s source-monitoring ability begins to develop at around the age of 4 or 5 years and continues to improve until the age of about 9 or 10 years (Roberts, 2002; Robinson & Whit- combe, 2003; Ruffman, Rustin, Garnham, & Parkin, 2001), it is possible that the younger children in our sample experienced more difficulty in distinguishing between what they knew on the basis of family stories and photographs and what they actually remem- bered. In contrast to this view, however, we found that there was no difference in either the age or content of participants’ earliest memories as a function of whether or not there were family photos or stories available to them.
Finally, the pattern of results found here may reflect age-related changes in the threshold for what individuals consider to be a “memory.” That is, adults may have a higher threshold than do children when identifying and reporting a memory of their past experiences. The finding that adults reported more details than did children or adolescents could be used to argue that adults set the bar higher, selecting older memories for which they can recall more detail. In contrast, the younger participants may have low- ered the bar, reporting earlier fragments of memories for which they could recall only a few details.
Individual differences in the threshold for identifying and re- porting early memories have received virtually no experimental attention to date despite the fact that individual differences in threshold have the potential to influence our estimates of child-
hood amnesia, even in studies conducted exclusively with adults. For example, a number of researchers have studied cultural and gender differences in individuals’ earliest memories. MacDonald et al. (2000), for example, compared the age of earliest memories of New Zealand European adults, New Zealand Maori adults, and Asian adults. They found that the Maori adults reported younger earliest memories than did members of the other two groups. Similarly, Mullen (1994) found gender-related and culture-related differences in the age of earliest memories. She found that female participants reported younger earliest memories than did male participants and that Caucasian adults reported younger earliest memories than did Asian adults. Although these data have tradi- tionally been interpreted as gender- and culture-related differences in the boundary of childhood amnesia (similar to the interpretation of the data reported here), they might also reflect gender-related and culture-related differences in the threshold that participants use to select a memory in the first place. Even if this is the case, however, some kernel of information about an individual’s very early experiences remains available and, under some conditions, may still be accessible to recall. Changing the instructions we use to get participants to retrieve their earliest recollections may shed new light on this issue.
In conclusion, the finding that the boundary and density of childhood amnesia change over the course of human development is consistent with the explanation that young children encode early events poorly and that the traces of such events are more rapidly forgotten and are less accessible to retrieval after very long delays. In this way, age-related changes in basic memory mechanisms may make an important contribution to the phenomenon of childhood amnesia. This is not to say that other factors do not play a role as well. The data reported here underscore the value of a develop- mental perspective in understanding this pervasive memory phe- nomenon.
References
Bachevalier, J. (1992). Cortical versus limbic immaturity: Relationship to infantile amnesia. In M. Gunnar & C. Nelson (Eds.), Developmental behavioral neuroscience (pp. 129–153). Hillsdale, NJ: Erlbaum.
Bauer, P. (2007). Remembering the times of our lives: Memory in infancy and beyond. Mahwah, NJ: Erlbaum.
Brainerd, C., & Reyna, V. (1990). Gist is the grist: Fuzzy-trace theory and the new intuitionism. Developmental Review, 10, 3–47.
Buckner, J. P., & Fivush, R. (1998). Gender and self in children’s auto- biographical narratives. Applied Cognitive Psychology, 12, 407–429.
Butler, S., Gross, J., & Hayne, H. (1995). The effect of drawing on memory performance in young children. Developmental Psychology, 31, 597– 608.
Cheatham, C. L., & Bauer, P. J. (2005). Construction of a more coherent story: Prior verbal recall predicts later verbal accessibility of early memories. Memory, 13, 516–532.
Cleveland, E. S., & Reese, E. (2008). Children remember early childhood: Long-term recall across the offset of childhood amnesia. Applied Cog- nitive Psychology, 22, 127–142.
Cohen, J. (1973). Eta-squared and partial eta-squared in fixed factor ANOVA designs. Educational and Psychological Measurement, 33, 107–112.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Cramér, H. (1999). Mathematical methods of statistics. Princeton, NJ: Princeton University Press.
1059AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Crovitz, H., & Harvey, M. (1979). Early childhood amnesia: A quantitative study with implications for the study of retrograde amnesia after brain injury. Cortex, 15, 331–335.
Crovitz, H., Harvey, M., & McKee, D. (1980). Selecting retrieval cues for early-childhood amnesia: Implications for the study of shrinking retro- grade amnesia. Cortex, 16, 305–310.
Crovitz, H., & Quina-Holland, K. (1976). Proportion of episodic memories from early childhood by years of age. Bulletin of the Psychonomic Society, 7, 61–62.
Dudycha, G., & Dudycha, M. (1933a). Adolescents’ memories of pre- school experiences. Journal of Genetic Psychology, 42, 468–480.
Dudycha, G., & Dudycha, M. (1933b). Some factors and characteristics of childhood memories. Child Development, 4, 265–278.
Fitzgerald, J. (1991). A developmental account of early childhood amnesia. Journal of Genetic Psychology, 152, 159–171.
Fivush, R., Gray, J., & Fromhoff, F. (1987). Two-year-olds talk about the past. Cognitive Development, 2, 393–409.
Fivush, R., & Hamond, N. (1990). Autobiographical memory across the preschool years: Toward reconceptualizing childhood amnesia. In R. Fivush & J. Hudson (Eds.), Knowing and remembering in young chil- dren (pp. 223–248). New York, NY: Cambridge University Press.
Fivush, R., Hamond, N., Harsch, N., Singer, N., & Wolf, A. (1991). Content and consistency in young children’s autobiographical recall. Discourse Processes, 14, 373–388.
Fivush, R., & Nelson, K. (2004). Culture and language in the emergence of autobiographical memory. Psychological Science, 15, 573–577.
Fivush, R., & Schwarzmueller, A. (1998). Children remember childhood: Implications for childhood amnesia. Applied Cognitive Psychology, 12, 455–473.
Freud, S. (1963). Introductory lectures on psycho-analysis. In J. Strachey (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 15, pp. 199–201). London, England: Hogarth Press. (Original work published 1905)
Gross, J., & Hayne, H. (1998). Drawing facilitates children’s verbal reports of emotionally laden events. Journal of Experimental Psychology: Ap- plied, 4, 163–179.
Hayne, H. (2004). Infant memory development: Implications for childhood amnesia. Developmental Review, 24, 33–73.
Hayne, H., Garry, M., & Loftus, E. F. (2006). On the continuing lack of scientific evidence for repression. Behavioral and Brain Sciences, 29, 521–522.
Henri, V., & Henri, C. (1898). Earliest recollections. Popular Science Monthly, 21, 108–115.
Howe, M. (2000). The fate of early memories: Developmental science and the retention of childhood experiences. Washington, DC: American Psychological Association.
Howe, M. L., & Courage, M. L. (1993). On resolving the enigma of infantile amnesia. Psychological Bulletin, 113, 305–326.
Howe, M. L., & Courage, M. L. (1997). The emergence and early devel- opment of autobiographical memory. Psychological Review, 104, 499– 523.
Howe, M. L., Courage, M. L., & Peterson, C. (1994). How can I remember when “I” wasn’t there: Long-term retention of traumatic experiences and emergence of the cognitive self. Consciousness and Cognition: An International Journal, 3, 327–355.
Howes, M., Siegel, M., & Brown, F. (1993). Early childhood memories: Accuracy and affect. Cognition, 47, 95–119.
Jack, F., & Hayne, H. (2007). Eliciting adults’ earliest memories: Does it matter how we ask the question? Memory, 15, 647–663.
Jack, F., MacDonald, S., Reese, E., & Hayne, H. (2009). Maternal remi- niscing style during early childhood predicts the age of adolescents’ earliest memories. Child Development, 80, 496–505.
Johnson, M., Hashtroudi, S., & Lindsay, D. (1993). Source monitoring. Psychological Bulletin, 114, 3–28.
Johnson, M., & Raye, C. (1981). Reality monitoring. Psychological Re- view, 88, 67–85.
Kihlstrom, J., & Harackiewicz, J. (1982). The earliest recollection: A new survey. Journal of Personality, 50, 134–148.
MacDonald, S., & Hayne, H. (1996). Child-initiated conversations about the past and memory performance by preschoolers. Cognitive Develop- ment, 11, 421–442.
MacDonald, S., Uesiliana, K., & Hayne, H. (2000). Cross-cultural and gender differences in childhood amnesia. Memory, 8, 365–376.
McCabe, A., Capron, E., & Peterson, C. (1991). The voice of experience: The recall of early childhood and adolescent memories by young adults. In A. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 137–173). Hillsdale, NJ: Erlbaum.
Morrison, C., & Conway, M. (2010). First words and first memories. Cognition, 116, 23–32.
Mullen, M. (1994). Earliest recollections of childhood: A demographic analysis. Cognition, 52, 55–79.
Nadel, L., & Zola-Morgan, S. (1984). Infantile amnesia: A neurobiological perspective. In M. Moscovitch (Ed.), Infant memory: Its relation to normal and pathological memory in humans and other animals (pp. 145–172). New York, NY: Plenum Press.
Neisser, U. (1962). Cultural and cognitive discontinuity. In T. E. Gladwin & W. C. Sturtevant (Eds.), Anthropology and human behavior (pp. 54–71). Washington, DC: Anthropological Society of Washington.
Nelson, C. A. (1998). The nature of early memory. Preventive Medicine: An International Journal Devoted to Practice and Theory, 27, 172–179.
Nelson, K. (1989). Monologue as representation of real-life experience. In K. Nelson (Ed.), Narratives from the crib (pp. 27–72). Cambridge, MA: Harvard University Press.
Nelson, K. (1992). Emergence of autobiographical memory. Human De- velopment, 35, 172–177.
Nelson, K. (1996). Language in cognitive development: The emergence of the mediated mind. New York, NY: Cambridge University Press.
Nelson, K. (2003). Narrative and self, myth and memory: Emergence of the cultural self. In R. Fivush & C. Haden (Eds.), Autobiographical memory and the construction of a narrative self: Developmental and cultural perspectives (pp. 3–28). Mahwah, NJ: Erlbaum.
Nelson, K., & Fivush, R. (2004). The emergence of autobiographical memory: A social cultural developmental theory. Psychological Review, 111, 486–511.
Newcombe, N., Lloyd, M., & Ratliff, K. (2007). Development of episodic and autobiographical memory: A cognitive neuroscience perspective. In R. V. Kail (Ed.), Advances in child development and behavior (Vol. 35, pp 37–85). San Diego, CA: Elsevier.
Perner, J., & Ruffman, T. (1995). Episodic memory and autonoetic con- sciousness: Developmental evidence and a theory of childhood amnesia. Journal of Experimental Child Psychology, 59, 516–548.
Peterson, C. (1994). Narrative skills and social class. Canadian Journal of Education, 19, 251–269.
Peterson, C. (2002). Children’s long-term memory for autobiographical events. Developmental Review, 22, 370–402.
Peterson, C., Grant, V., & Boland, L. (2005). Childhood amnesia in children and adolescents: Their earliest memories. Memory, 13, 622– 637.
Peterson, C., & Rideout, R. (1998). Memory for medical emergencies experienced by 1- and 2-year-olds. Developmental Psychology, 34, 1059–1072.
Peterson, C., Wang, Q., & Hou, Y. (2009). “When I was little”: Childhood recollections in Chinese and European Canadian grade school children. Child Development, 80, 506–518.
Pillemer, D., & White, S. (1989). Childhood events recalled by children and adults. Advances in Child Development and Behavior, 21, 297–340.
Reyna, V., & Brainerd, C. (1995). Fuzzy-trace theory: An interim synthe- sis. Learning and Individual Differences, 7, 1–75.
1060 TUSTIN AND HAYNE
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Richardson, R., & Hayne, H. (2007). You can’t take it with you: The translation of memory across development. Current Directions in Psy- chological Science, 16, 223–227.
Roberts, K. (2002). Children’s ability to distinguish between memories from multiple sources: Implications for the quality and accuracy of eyewitness statements. Developmental Review, 22, 403–435.
Robinson, E., & Whitcombe, E. (2003). Children’s suggestibility in rela- tion to their understanding about sources of knowledge. Child Develop- ment, 74, 48–62.
Rubin, D. (1982). On the retention function for autobiographical memory. Journal of Verbal Learning and Verbal Behavior, 21, 21–38.
Rubin, D. (2000). The distribution of early childhood memories. Memory, 8, 265–269.
Ruffman, T., Rustin, C., Garnham, W., & Parkin, A. (2001). Source monitoring and false memories in children: Relation to certainty and executive functioning. Journal of Experimental Child Psychology, 80, 95–111.
Schachtel, E. G. (1947). On memory and childhood amnesia. Psychiatry, 10, 1–26.
Sheingold, K., & Tenney, Y. (1982). Memory for a salient childhood event. In U. Neisser (Ed.), Memory observed (pp. 201–212). San Francisco, CA: Freeman.
Simcock, G., & Hayne, H. (2002). Breaking the barrier? Children fail to translate their preverbal memories into language. Psychological Science, 13, 225–231.
Simcock, G., & Hayne, H. (2003). Age-related changes in verbal and nonverbal memory during early childhood. Developmental Psychology, 39, 805–814.
Suddendorf, T., & Corballis, M. (1997). Mental time travel and the evo- lution of the human mind. Genetic, Social, and General Psychology Monographs, 123, 133–167.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson (Eds.), Organization of memory (pp. 381–403). New York, NY: Academic Press.
Tulving, E. (2002a). Chronesthesia: Conscious awareness of subjective time. In D. Stuss & R. Knight (Eds.), Principles of frontal lobe function (pp. 311–325). New York, NY: Oxford University Press.
Tulving, E. (2002b). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 1–25.
Van Abbema, D., & Bauer, P. (2005). Autobiographical memory in middle childhood: Recollections of the recent and distant past. Memory, 13, 829–845.
Waldfogel, S. (1948). The frequency and affective character of childhood memories. Psychological Monographs, 62, 1–39.
Wang, Q. (2001). Cultural effects on adults’ earliest childhood recollection and self-description: Implications for the relation between memory and the self. Journal of Personality and Social Psychology, 81, 220–233.
Wang, Q. (2003). Infantile amnesia reconsidered: A cross-cultural analysis. Memory, 11, 65–80.
Wang, Q., Conway, M., & Hou, Y. (2004). Infantile amnesia: A cross- cultural investigation. Cognitive Sciences, 1, 123–135.
Welch-Ross, M. (2000). Personalizing the temporally extended self: Eval- uative self-awareness and the development of autobiographical memory. In C. Moore & K. Lemmon (Eds.), The self in time: Developmental perspective (pp. 97–120). Mahwah, NJ: Erlbaum.
West, T., & Bauer, P. (1999). Assumptions of infantile amnesia: Are there differences between early and later memories? Memory, 3, 257–278.
Wetzler, S., & Sweeney, J. (1986). Childhood amnesia: An empirical demonstration. In D. Rubin (Ed.), Autobiographical memory (pp. 191– 201). New York, NY: Cambridge University Press.
Wright, F. (2006). Childhood amnesia: Retrospective studies, prospective studies, and theoretical explanations (Unpublished doctoral disserta- tion). University of Otago, Dunedin, New Zealand.
Received April 12, 2007 Revision received April 26, 2010
Accepted April 27, 2010 �
1061AGE-RELATED CHANGES IN CHILDHOOD AMNESIA
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
