Previous research on the misinformation effect has demonstrated that memory for events can be distorted by suggestive information presented after the fact. A separate line of research on choice blindness has shown that after making a choice between multiple alternatives, if people are told they chose something different from what they actually chose, they often fail to notice the discrepancy. In a methodological marriage of these two paradigms, participants first witnessed an event, and were then asked ten memory questions about what they saw. Later, their responses were presented back to them for review; half of the participants were told that these responses were their own reports from earlier, and half were told that they were responses from another subject. Importantly, three of the ten responses shown during this review were systematically altered from what the participants had originally said. At a final test, participants answered the initial memory questions a second time. Repeated measures ANOVAs revealed that viewing manipulated reports caused participants' memories to change in ways consistent with the misinformation. This effect was stronger when the misinformation was presented as "what another participant had said" than when it was presented as "what you previously said." This "memory-blindness" finding represents an important avenue for future research on the accuracy of eyewitness reports.

People lie about the past. They may lie in order to avoid consequences, to facilitate social bonding, or to make sense of their knowledge and their experiences. These reasons closely parallel proposed functions of memory flexibility, and a handful of studies have shown that under certain conditions, fabricating past events can lead to false memories and beliefs. Little research, however, has specifically examined whether lying about the personal past can alter autobiographical memory. The present studies are an investigation of how fabricating past experiences influences autobiographical memory, and the role of self-enhancing biases in guiding this influence.

Across three studies, participants were prompted to compose written fabrications about a childhood event that was unlikely to have occurred. Specifically, they wrote stories were rich with perceptual and emotional detail and embellished with plausible false content. After a short delay, they were asked to disregard their fabrications and provide sincere memory reports. Results indicated that fabrications influenced the subsequent memory reports, especially when the fabrications were thematically self-enhancing. Study 2 was a replication of Study 1, with additional experimental conditions intended to disentangle the influence of emotional valence from the self-enhancing features of the events, and to address the possibility that changes in mood accounted for the observed patterns. Results indicated that neither event valence, nor mood, could explain the patterns observed in Studies 1 and 2. In a third and final study, participants first completed one of three possible self-reflection tasks, which varied in their self-affirming and self-threatening characteristics. Next, they completed a fabrication task and memory questionnaire, as in Studies 1 and 2. Results indicated that self-affirming reflections yielded patterns of memory distortion similar to Studies 1 and 2, while self-threatening reflections appeared to reverse that pattern--threatened participants were less likely to incorporate self-enhancing fabrications into memory, and more likely to incorporate self-diminishing fabrications. The findings are considered in the context of theoretical questions surrounding motivational influences on how people recall past events.

This dissertation addresses two basic questions: 1. Are people with highly superior autobiographical memory (HSAM) susceptible to memory distortions? 2. What is different about them that might offer clues that would help explain their ability? To answer the first question thoroughly, HSAM individuals and age match controls participated in a number of memory distortion tasks. In the DRM memory distortion word list paradigm we found that HSAM participants had comparably high rates of critical lure endorsement, indicating a vulnerability to false memories brought about by associations. They also participated in a classic misinformation experiment with photographic slides as the original event, and text narratives containing some pieces of misinformation. At the subsequent memory test HSAM individuals indicated more false memories than control participants, a finding that became non-significant when adjusting for individual differences in absorption. After a subsequent source test, HSAM and control participants had comparable numbers of false memories from misinformation.

In semi-autobiographical memory distortion tasks, HSAM and control participants had fairly similar rates overall. For example, in a nonexistent news footage task using suggestion (also known as the "crashing memory" paradigm) 10% of HSAM individuals said they had seen the footage (a further 10% indicated maybe/unsure), whilst 18% of controls did (5% maybe; ns). A guided imagery task, with the same nonexistent footage as the target event, produced similarly increased rates of false report in HSAM (17% changed from "no" to "yes") and control (10% from "no" to "yes") participants. Memory for their emotions in the week after 9/11 was similarly inconsistent in HSAM and control participants. These results suggest that, relative to controls, HSAM individuals are as susceptible to both misinformation and reappraisals when the target events are semi-autobiographical.

The second main research question asked what is different about HSAM individuals that might give us clues as to why they have their ability? To answer this we measured HSAM participants' and age/gender matched controls' on a number of behavioral measures to test three main hypotheses: imaginative absorption, emotional arousal, and sleep. HSAM participants were significantly higher than controls on two dispositions--absorption and fantasy proneness. These two dispositions were associated with a measure of HSAM ability within the superior memory participants. The emotional arousal hypothesis yielded only weak support. The sleep hypothesis was not supported in terms of quantity, but sleep quality may be a small factor worthy of further research. Other individual differences are also documented. Speculative pathways describing how absorption and fantasizing could lead to enhanced autobiographical memory are discussed.