Memory Distortions Induced by Fluency ? Behavioural and Electrophysiological Investigations

Detta är en avhandling från Department of Psychology, Box 213, 22100 Lund

Sammanfattning: The investigation of memory distortions provides insights into how memory normally works. Adopting a cognitive neuroscience perspective, the present thesis aimed at investigating memory distortions and the neural correlates of processes contributing to successful control of memory interference in healthy participants. All studies employed variations of the modified continuous recognition task. In the first run of this task, participants are presented with a series of items. Some pictures are repeatedly presented as targets and participants have to indicate recurrences. In the second run, the same series of pictures is presented again, but target-distracter status is changed. Participants again have to detect recurrences of targets, but only in the ongoing run. The first run assesses discrimination of old and new information, whereas the second run measures the ability to handle interference from irrelevant memory traces (i.e., distracters that are familiar from the previous run). In order to induce additional interference, factors such as emotion and attractiveness were used as these item characteristics were expected to enhance processing fluency. Paper I investigates whether memory distortions can be elicited when participants have to manage interference from irrelevant emotional memories. Paper II examines whether memory distortions can also be induced for neutral faces when targets and distracters are highly similar. To this end, four continuous recognition runs instead of two were employed. Papers I and II demonstrate that emotionally valenced faces induce distortions, both during old-new recognition in the first run and when participants had to distinguish between previously seen targets and distracters in the subsequent runs. These findings are consistent with the idea that emotion induces errors when the salience of emotional information is considered diagnostic for memory judgments. Interestingly, in Paper II memory distortions were also induced for neutral faces after four runs, when targets and distracters were highly similar. This finding suggests that memory interference can be induced when the salience of irrelevant distracters is high, that is, it can be caused by either emotional or mnemonic salience. Paper III uses event-related potentials(ERPs) to investigate memory control functions supporting successful control of memory interference and examine whether those processes are modulated by emotion. Capitalising on ERP memory effects, the results suggest that emotion modulates the extent to which recollection was employed to ensure successful target-distracter differentiation in the subsequent runs. Recollection contributed to the recognition of emotional and neutral target faces as well as to the rejection of negative distracters. Thus, recollection was used to correctly reject irrelevant negative memories. Paper IV employs famous and unfamiliar faces varying in attractiveness in order to explore (a) whether facial attractiveness induces memory distortions in a similar fashion as facial expressions do and (b) whether distinctive semantic knowledge associated with famous faces may reduce such distortions. The results demonstrate that attractiveness induced interference for unfamiliar faces in the first run and for famous faces in the second run. Although distinct semantic knowledge may prevent distortions for famous faces during old-new recognition in the first run, it seems inefficient under challenging conditions in the subsequent run. The results of the four studies are discussed within dual-process models of recognition memory, the source-monitoring framework, as well as cognitive neuroscience models of memory control. In line with previous proposals it is argued that principles like the fluency heuristic supporting true recognition, can under some circumstances also cause false recognition. In this vein, memory distortions cannot be regarded as system errors but rather as consequences of the reconstructive nature of memory, illustrating that remembering is an inferential and attributional function.

  Denna avhandling är EVENTUELLT nedladdningsbar som PDF. Kolla denna länk för att se om den går att ladda ner.