This thesis investigated the characteristics of childhood memories that remain accessible over the whole life span. On the assumption that one of the primary purposes of autobiographical memory is its adaptive function, it was hypothesised that, in order to fulfil this function, autobiographical memories must include an affective component. That is, each memory will consist of a record of an experienced event together with the accompanying emotion so as to mediate an appropriate behavioural response in similar future circumstances. For the same reason, it was also predicted that, as a person grows older, selected childhood memories will still retain their emotionality, vividness and personal importance. A detailed analysis of the nature of childhood memories was undertaken and evidence to support these hypotheses was sought in three separate studies. In Study 1, a structured interview was used and 109 people aged between 21 and 91 years were asked to free recall memories drawn from within three - year periods of childhood from birth to 18 years. Subjects' ratings were used to explore the characteristics of these memories and the effects of age encoding subject age and subject gender were systematically examined. In contrast with previous research, it was found that 83 % of subjects could recall memories from below three years of age. Uniformly high ratings of emotionality showed that the affective component, as predicted, was high across all encoding. ages and showed no decline with subject age. Ratings of clarity and realism were also unaffected by subject age. The incidence and vividness of sensory imagery increased with age of encoding, but, again, did not decline with subject age. Although most vivid memories involved imagery, 15%of subjects sometimes claimed to recall vividly without imaging. These findings suggest that the affective component may have been contributing to overall vividness. Study 2 using, cued recall, compared latency of response for emotion cued words and object cued words and also used ratings to examine the characteristics of the retrieved memories. Memories were retrieved Faster to object cues but these memories were still rated high in emotionality. Memories retrieved to emotion cues, although slower to access, were rated as more important, more emotional, more unusual and more frequently recalled. Cue type was therefore shown to be a powerful factor in determining accessibility, overriding other memory characteristics. Although in Studies 1 and 2 subjects selected their own memories, Study 3 tested the retention of details of early experience of school life as designated by the researcher. It was found that detailed memories could still be recalled even in old a`e and that the rate of forgetting declined steadily with age This evidence of persisting retention of early childhood experience is consistent with the view that such memories serve a functional role in that they are an integral part of the 3 individual's self history and developing self concept. The research provides a substantial body of data detailing, the topics recalled From different periods of childhood and the nature of the memories and charts the remarkable stability of these memories across the life span.

No / In a memory survey, adult respondents recalled, dated, and described two earliest positive and negative memories that they were highly confident were memories. They then answered a series of questions that focused on memory details such as clothing, duration, weather, and so on. Few differences were found between positive and negative memories, which on average had 4/5 details and dated to the age of 6/6.5 years. Memory for details about activity, location, and who was present was good; memory for all other details was poorer or at floor. Taken together, these findings indicate that (full) earliest memories may be considerably later than previously thought and that they rarely contain the sort of specific details targeted by professional investigators. The resulting normative profile of memory details reported here can be used to evaluate overly specific childhood autobiographical memories and to identify memory details with a low probability of recall.

Yes / In a large-scale survey, 6,641 respondents provided descriptions of their first memory and their age when they encoded that memory, and they completed various memory judgments and ratings. In good agreement with many other studies, where mean age at encoding of earliest memories is usually found to fall somewhere in the first half of the 3rd year of life, the mean age at encoding here was 3.2 years. The established view is that the distribution around mean age at encoding is truncated, with very few or no memories dating to the preverbal period, that is, below about 2 years of age. However, we found that 2,487 first memories (nearly 40% of the entire sample) dated to an age at encoding of 2 years and younger, with 893 dating to 1 year and younger. We discuss how such improbable, fictional first memories could have arisen and contrast them with more probable first memories, those with an age at encoding of 3 years and older.