Social and affective influences on memory accuracy
An initial survey of the history of the psychology of memory indicated that while research interest has been intense, remarkably little insight has been gained into the nature of human memory. While a number of methodological and empirical difficulties in the investigation of memory are identified, it is concluded that the bulk of research has been insufficiently directed towards memory for real-world events. It is suggested that as a consequence, this has led to sterile debates on issues which are of dubious significance. In a bid to circumnavigate these problems, the thesis attempts to investigate the 'psychological' aspects of memory. More specifically, these comprise the higher order memory processes which are required to manage and support social perception and memory. These include any aspects of social affect and cognition which are conceptually-driven and based on extra-stimulus information (for example, inferences, attributions, beliefs, motives, desires, stereotypes, contextual knowledge). Three memory frameworks are reviewed - Schematic (Bartlett 1932), Headed Record (Morton, Hammersley, and Bekerian 1981, 1985) and Multiple Entry (Johnson 1983) - each of which is capable of generating accounts of memory for real-world events. The provision of this theoretical context was aimed at (a) interpreting the data from the Experiments, and (b) assessing which of the three appears to be of the greatest utility. In addition, a selective review of some of the ways social psychologists have characterized social cognition is included, with the aims of: (a) highlighting the need for such cognition to be active, constructive, and inferential, and (b) to articulate further the requirements for which the three memory frameworks under consideration must account. Since the aspects of memory under review in this thesis are principally those which are not objectively present in the stimulus, but instead constructed by the subject, a potential source of inaccuracy exists for the perception and memory of real-world events. Accordingly, in addition to their theoretical significance, the data were assessed in the light of their potential practical ramifications for the reliability of witness testimony. Experiments 1 to 4 examine the action of racial prejudice on memory for a real-world event. After establishing that a racial stereotype exists in Experiment 1, Experiment 2 examined its potential influence on subjects' memory for a verbal, eyewitness account of an event, while Experiments 3 and 4 assessed its influence on the subjects' own eyewitness accounts for the same, visually presented event. It is concluded that memory could be significantly influenced by higher order processes which are driven by stereotypical expectations. However, the effects were only observed under certain conditions and it is suggested that the influence of racial prejudice is not as pervasive as has often been claimed. Experiment 5 not only investigated the impact of extra-stimulus information on subjects' memory for a real-world event but also attempted to isolate the locus of its influence in information processing. By employing a between groups design which provided subjects with contextual information concerning a central character's occupation (variously: prior to the event, immediately after the event, or immediately before the recognition test a week later), it was hoped to isolate the locus and/or loci of its action. It is concluded from this study that the extra-stimulus information influenced memory only (a) where it was introduced adjacent to encoding the stimulus, and (b) where sufficient time for rehearsal had been permitted. Experiments 6 and 7 used real-world stimulus material taken from football matches played between England and Scotland, and, Celtic and Rangers, respectively. In both cases, the experimental interest was to attempt to define the locus or loci of higher order activity in information processing, by examini.