Bruce, D., Dolan, A. & Phillips-Grant, K. (2000). On the transition from childhood amnesia to the recall of personal memories. Psychological Science, 11, 360-364.
Childhood amnesia is the impoverished recall by adults of autobiographical memories from early childhood. This type of memory may also be referred to as episodic memory (Tulving), personal or recollective memory (Brewer)
The focus of this research is to estimate when the transition from childhood amnesia to recollection of personal memories occurs.
Other research has looked at both childhood specific memories and also retention of autobiographical memories over the life-span.
Childhood amnesia may be studied through a number of self-reports
*Describe & date earliest personal memories
*Report a number of experiences prior to a certain age.
*Describe experiences from childhood then date when they happened
*Describe known events that occurred at various ages.
Results show that adults earliest memories of childhood tend to take place around 3 or 4 years of age.
Wetzler & Sweeney have found that college students have a poorer memory for events before the age of 8 than for events taking place after this age. Their research shows that childhood amnesia is common up to age 5. Other research suggests that childhood amnesia may extend up to age 6 (Winograd & Killinger).
Conclusion: The transition from childhood amnesia to episodic memory occurs somewhere between 3 & 7 years of age. To define the age more specifically, we may look at the transition in terms of an absolute threshold. The threshold is the age beyond which 50% of the remembered events occurred. This just means that the threshold is the median age. The median age may be used as an estimate of when episodic memory begin to "outweigh" childhood amnesia.
Although the median age may be used as an estimate, it is a biased measurement because it only approached the threshold from subjects' personal memories. A better estimate would be to combine the above approach with an approach that estimated the median age for amnesiac events. This combined estimate would be absent of a directional bias.
The goal of the present study was to produce these two distributions.
The memory distribution was created by having subjects describe childhood events that they remembered happening and to date them.
The amnesiac distribution was created by having subjects describe and date childhood events that
they knew happened but did not remember.
An estimate of the transition, the threshold, is obtained by averaging the median ages of the remember and know distributions.
Method
Participants (N=133) were asked to describe two events that occurred prior to the age of 8, one event that they remembered and one event that they knew happened. Participants were tested individually or in pairs and report order was averaged across sessions. Subjects wrote brief descriptions of the remember/know events in test booklets. They recorded their approximate age and gave a confidence judgment of this estimate.
After 76 subjects granted permission, letters were sent to people (usually parents) who might authenticate the events. The data of 18 subjects were excluded for various reasons. Data of 95 subjects, including 70 female and 25 male, were used in analysis.
Results
The mean age for know events, 3.35 (SD+1.86), is significantly younger than the mean age for remember events, 5.85(SD+1.49);p<.001.
Neither the order of events nor the type of event had a significant effect on the reported age of events. However, when know events were reported second, participants were less confident in their age estimates for the know events than for the remember events.
Of the letters sent of 74% were returned; this is 58% relative to the total number of subjects whose data was used.
Parents largely agreed with event descriptions that their children provided. Out of 55, parents judged 49 to have happened as occurred, 2 to have occurred save for minor differences in details, and they had no knowledge of the 4 remaining.
There was no significance between the estimation of the parents or the children for the age that the events were to have occurred.
These events were further analyzed across four categories: topic, location, affect, and people involved. A greater proportion of remember events occurred outside--.44 versus .23. Know events never contained direct descriptions of affect whereas .34 of remember events did.
The data of the present study closely corresponds with Waldfogel's research of early memories. See Fig. 1.
The distribution for the remember and know events is represented in Fig. 2. The median age for the know distribution is 3.2 and the median age for the remember distribution is 6.07. The mean of these two estimations is 4.64 which can be concluded as the estimate of when childhood amnesia begins to be outweighed by the retrievability of personal memories.
Discussion
The current study has taken a new, two sided approach to estimating the transition time between childhood amnesia and episodic memory. The results suggest that previous studies have overestimated the duration of childhood amnesia.
Possible problems in the current study:
*Events that parents want their children to know about happened at an early age
in the child's life. Bruce et al argues that noteworthy events are equally likely to occur at any age and the events that the children do not remember are due to childhood amnesia.
*There may be problems with the implications of the psychophysical method used
in the current study. Bruce et al recognizes this but argues that the usefulness of
the heuristic outweighs any possible problems. The heuristic is useful because it
prompts examination of childhood amnesia from two sides.
*4.6 is not a concrete transition age for every person. It is only an average which
may vary depending on a number of circumstances such as type of event, age of
subject, birth order, gender, and cultural group.
Onset of autobiographical memories is the first retrievable childhood memory. This is not what the current study is concerned with. The current study examines the Transition between childhood amnesia and autobiographical memory. The average transition age, 4.64 years, is when autobiographical memories tend to outweigh amnesia. This does not imply that children are in a transitional state where some things are remembered and some things are not.
The current study is compatible with many studies including a study on language, social, and narrative skills (Fivush & Schwarzmueller, 1998: Nelson, 1993; Pillemer, 1992), encoding schemes (Schachtel, 1947; Winograd & Killinger, 1983), and a study on person memory systems (Pillemer & White, 1989). The current study is also consistent with research of Perner & Ruggman that indicates that the ability to accurately remember events develops between 3 to 6 years of age.