Individual differences in reasoning have been observed in a wide variety of tasks. Descriptions of the variation in response have been framed in terms of use of different strategies that invoke different representations. This thesis argues that in order to convert descriptions into <i>explanations </i>of performance it is necessary to compare and combine psychometric accounts with computational accounts of the processes underlying representation selection and use. Descriptions of strategies, representations and algorithms and their inter-relationships are necessary for a full account of reasoning behaviour. Two large-scale studies of deductive reasoning are presented to illustrate this approach in action, and the inadequacy of accounts that do not provide accounts at all these levels. The first compares two theoretically motivated methods for solving categorical syllogisms, the second study assesses learning from and learning within a multimodal logic course called Hyperproof. These studies are compared to measures of spatial ability, field-independence/dependence, and serial/holist learning style. The interaction of students' styles of learning with different presentations of information generalises across the domains. This generality is best expressed when psychometric and computational accounts of reasoning are consolidated.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:657929 |
| Date | January 2000 |
| Creators | Monaghan, Padraic |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/15418 |
Page generated in 0.0021 seconds