Adult attachment and phenomenological characteristics of autobiographical memory

Abstract

The aim of the studies undertaken for this thesis was to explore relations between adult attachment and autobiographical memory. Study One investigated how a self-report measure of adult attachment style related to young adults' (N = 211) recall of their earliest memories. Dismissing individuals reported fewer negatively valenced memories than their counterparts in the secure and preoccupied groups. No attachment-related differences were found in the total number of memories (positive, neutral, negative) recalled, or individuals' ratings of the phenomenological properties of the memories. All three groups tended to rate negative memories more highly than neutral/positive memories on the phenomenological characteristics, although preoccupied individuals tended to show least differentiation on the basis of emotional valence. Study Two investigated how attachment state of mind as assessed using the Adult Attachment Interview (George, Kaplan, & Main, 1985) related to autobiographical memory in a separate sample (N = 65) of young adults. Autobiographical memory was assessed in terms of recall (a) of one's earliest memory, and (b) of childhood memories in response to attachment-related and non- attachment cues, and this study also controlled for concurrent depressive symptoms and previous experience of trauma. As in Study One, the earliest memory and the cued memories were rated for their phenomenological properties, but data were also collected on latency of recall. No relation was found between A.A.I, classification and any characteristic of the earliest memory. For the cued recall of attachment-related memories, A.A.I, classification independently predicted vividness, emotional intensity at encoding and emotional intensity at recall, with dismissing individuals scoring lowest and preoccupied highest. A.A.I, classification also predicted certain aspects of recall for non-attachment material. In particular, dismissing individuals rated non- attachment memories as less specific and less vivid than did individuals in the secure and preoccupied groups. A.A.I, classification has little impact on individuals’ responses to the attachment-related and non-attachment memories. The only effect of A.A.I, classification was seen on ratings of specificity; somewhat surprisingly, dismissing individuals rated attachment memories as more specific than non- attachment memories, whereas secure and preoccupied individuals did not differ in their ratings of the two types of memory. Study Three investigated how A.A.I, classification related to imagined future events in response to attachment-related and non-attachment cues in the same sample of participants who had taken part in Study Two. Controlling for gender, depressive symptoms and previous trauma (as in Study Two), the results of Study Three showed that A.A.I, classification predicted the reported vividness and self-relevance of attachment-related imagined future events. Compared with secure and preoccupied individuals, those in the dismissing group reported that future attachment-related events were less vivid. There was also a marginally significant trend for dismissing individuals to rate attachment-related future events as less self-relevant. Comparing recall of previous past events with future imagined events, individuals across all A.A.I, categories were slowing at recounting future events than at recalling past events, and rated past events as more vivid and emotionally intense. However, it was future events that were rated as more self-relevant than past events. Study Three also found that there was greater concordance between ratings of past and future events with respect to specific phenomenological properties for insecure individuals than for secure individuals.