City Research Online

Abstract

This thesis investigates the validity of a view of memory retrieval that insists on the diagnostic relationship between a target memory and a retrieval cue. The widely accepted principle of encoding-retrieval match proposes that retrieval performance is related to the degree to which the cues at the time of retrieval are compatible with those at the time of encoding (Tulving, 1979, 1983). The effect of cue-overload, another well documented effect, stipulates that memory retrieval performance is expected to decrease as the number of potential targets in memory subsumed under a retrieval cue increases (Watkins & Watkins, 1975). With these two effects interacting, a variety of outcomes are possible. The memory as discrimination view investigated as part ofthis thesis suggests that factors such as cue-overload, distinctiveness, context, and potentially others contribute to a process of memory retrieval where diagnosticity is central (Nairne, 2002). Under such a view encoding-retrieval match cannot be relied upon to generate predictions. Under a diagnostic retrieval system, retrieval is most likely to be successful when a cue specifies a target in memory while excluding other potential candidates. Other memory as discrimination views do exist. However, within this thesis the above view is the only one being considered.

The studies presented as part of this thesis investigate the memory as discrimination hypothesis by studying the interaction between cue-overload and encoding-retrieval match. Empirical investigations are included in which proportion of correct responses is considered. More formal work is also included in which response time and probability of correct recall data are called upon. Chapter 1 reviews the relevant literature in the field, while chapters 2 and 3 report empirical investigations testing the predictions of the memory as discrimination hypothesis. In chapters 2 and 3, paired associate learning tasks were used. In such a task, participants are presented with pairs of words under the instruction that they will later be asked to remember one member of a given pair using the other member as a retrieval cue. Chapters 2 and 3 present the results of studies where the encoding context was manipulated to investigate the roles of encoding-retrieval match and cue-overload in retrieval. This was achieved by using contexts that were unique to a target or associated with multiple targets and then orthogonally manipulating encoding-retrieval match and cue-overload. The experiments reported in chapter 2 used interactive contexts in the form of spot the difference images; participants were asked to spot the difference between two images; each time they were successful, they were provided a pair of words to associate. Cue-overload was manipulated by varying the number of word pairs associated with a given spot the difference image. At test, the first word of each pair was presented and participants were required to provide the correct second item. The reinstatement or non-reinstatement of the background images at test was employed as a means of manipulating encoding-retrieval match. Using fonts as context, the experiments reported in chapter 3 varied the number of word pairs per list that shared a font, and manipulated the reinstatement of originally viewed fonts to further test the findings of chapter two. The general prediction of the memory as discrimination view across experiments in chapters 2 and 3 is that a reduction in the level of encoding-retrieval match between conditions may result in a detrimental, null, or beneficial effect on retrieval performance depending on the presence of a cueoverload effect.

Chapters 4 and 5 offer mathematical representations of cue-overload, encodingretrieval match, and possible interactions between the two in accordance with a memory . as discrimination view. The predictions from these formal representations are qualitatively compared to the data from experiments that manipulated encoding-retrieval match and cue-overload. Formal representations of encoding-retrieval match and cueoverload in chapter 4 predict the contributions of encoding-retrieval match and cueoverload in response time. A formal model of the probability of correct retrieval is offered in chapter 5. The essential feature of the model presented in chapter 5 is in its generation of predictions in which the effect of encoding-retrieval match can be reversed by changes in cue-overload. The present thesis concludes that the effect of encodingretrieval match is highly dependent on cue-overload in several ways and can be seen as unreliable - and further, that context cues may serve to facilitate this dependence.

Publication Type:	Thesis (Doctoral)
Subjects:	B Philosophy. Psychology. Religion > BF Psychology
Departments:	School of Health & Psychological Sciences > Psychology Doctoral Theses School of Health & Psychological Sciences > School of Health & Psychological Sciences Doctoral Theses

Export

Downloads