A model for the processing of sequential information
| dc.contributor.author | MacGregor, James Norman | |
| dc.contributor.supervisor | Bavelas, Alex | |
| dc.date.accessioned | 2025-06-17T00:20:41Z | |
| dc.date.available | 2025-06-17T00:20:41Z | |
| dc.date.issued | 1979 | |
| dc.degree.department | Department of Psychology | |
| dc.description.abstract | The dissertation proposed and tested a quantitative model of the processing of sequential information. In the first chapter, criticisms were made of the most generally accepted current quantification of sequential patterns. First, the argument was forwarded that the high correlations which support the measure are not truly representative of its predictive power. It was demonstrated that the measure typically takes on a highly limited range of values, and for this reason cannot account for much of the systematic variation found in behaviour. Second, it was argued that the measure is conceptually limited and cannot account for the results of a particular class of sequential processing studies concerned with temporal patterning. In the second chapter a new quantitative model was proposed, which provides a measure of sequential patterns called " structural embeddedness" or "embeddedness-of-runs." The measure is sensitive to the relative positions of the larger and smaller subjective structures within a sequential pattern. It takes on a minimum value when the magnitudes of structures vary inversely with their degree of embeddedness within a pattern. This means that the lowest value obtains when the largest subjective structures are in the outermost positions, the next largest structures in the next outermost positions, and so on. It takes on a maximum value when the opposite is true, and the largest structures are in the innermost positions, the next largest in the next innermost, and so on. Comparisons of the structural embeddedness of sequences with behavioural data in the literature indicated that subjects appear to prefer patterns with minimal structural embeddedness. Other results examined indicated that the subjective complexity of sequences varies directly with the degree of structural embeddedness. Two main hypotheses were proposed: that the complexity of sequences varies as a direct function of structural embeddedness, and that sequences tend to be organized into forms which minimize their structural embeddedness. In the third chapter psychological interpretations of the model’s two parameters were considered. The embeddedness of structures was proposed to reflect the order in which they are processed, with outermost structures being processed first, next outermost, next, and so on, resulting in an "endsinward" type of processing. The magnitude of structures was given two possible interpretations: either larger structures required more effort, in the process of forming or, once formed, they represented more salient features of a sequence. The order- of- processing hypothesis, and the effort or salience hypothesis were proposed as subsidiary hypotheses. The fourth chapter described the general method used in testing the experimental hypotheses. The method involved tracking subjects' anticipations during processing, and classifying responses into correct anticipation, wrong anticipation and no anticipation (copy response). The fifth chapter described two experiments. Both experiments found significant differences in the complexity of sequences across levels of structural embeddedness, under conditions where other features of sequences were controlled. This supported the first hypothesis. The second experiment demonstrated a significant tendency for sequences to be reorganized towards minimizing structural embeddedness. This result supported the second hypothesis. In addition, the first experiment confirmed that embeddedness reflects order-of-processing, and indicated that the magnitude of structures affects their salience, with larger structures being more readily processed. The structural embeddedness measure was therefore interpreted as reflecting the degree to which the most readily processed structures of a sequence are in the most readily processed positions. Such an arrangement appears to maximize subjective pattern "goodness." Hypotheses were proposed to explain this finding, and a more general psychological theory outlined. The final chapter considered possible extensions of the theory into the related areas of serial learning and verbal processing. Finally, an issue more central to the basic theory was discussed. | |
| dc.description.scholarlevel | Graduate | |
| dc.identifier.uri | https://hdl.handle.net/1828/22391 | |
| dc.language.iso | en | |
| dc.rights | Available to the World Wide Web | |
| dc.title | A model for the processing of sequential information | |
| dc.type | Thesis |