Ericksen, NeilKatterns, RobertBarbour, Robert H.2026-08-062026-08-061993https://hdl.handle.net/10289/18515Pask’s “Conversation Theory”, is an influential view of the nature of conversations between people, especially in learning situations. The theory provided a predictive framework within which to extend Pask’s laboratory work. Existing work used participant human observers to study the learning process. This technique may be unsuited to the normal classroom context. Software extensions to a geographical image library, map and data store, the BBC Domesday System, were used to extend Pask’s work. In carrying out this research, the three main principles from Conversation Theory were examined: 1) that people learn in different ways; 2) that the degree of matching between the learning strategies of teachers and learners strongly influences the rate of learning; and 3) that learning takes place as a conversation between participants whose essential features can be successfully automated. This thesis examined and evaluated three questions derived from Pask’s Conversation Theory in the light of empirical results collected in Phase 1 of a two phase research design. These questions were: 1) whether differences in learning preferences were detectable in modified versions of Ford’s (1985) questionnaire; 2) whether preferences for learning strategies were reflected in individual differences in learning behaviour; 3) whether matching learning preferences with learning experiences enhanced the learning process. Methods of investigation in Phase 1 included: a modified questionnaire; and using the BBC Domesday System to capture learning behaviour with a series of software tools. Findings show: that differences in learning preferences were detectable in questionnaire responses; that differences in learning behaviour were detected also in learning performances; that differences in preferences and behaviour were minimally correlated. Questions emerged from the lack of expected strength of agreement between findings of Phase 1. These questions are discussed in Phase 2 in relation to recent developments in both epistemology and the theory of computer science. The two further questions guiding the research in Phase 2 were: 4) whether matching learning and teaching strategies enhances learning; 5) whether a computer can act as a conversation participant in a teaching-learning context. Methods employed in Phase 2 include an examination of the findings of Phase 1 in the light of recent philosophical literature. Findings from Phase 2 show why matching is undecidable and why a computer cannot contribute to a conversation in the way a person does. Results in Phase 2 uphold Pask’s first principle, leave the second undecided and cast serious theoretical doubt on the third. The implications of the findings for educators, geographers and computer scientists are discussed.enAll items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.Thesis