An investigation of the effects of presentation mode and message format in the human-computer interface: A dual coding approach

Accounting information is the basis for business decisions made by various non-accountant users such as investors, creditors, and managers. Although each group has different information needs, accountants should provide information that is useful in making informed decisions.; To be useful, information must be communicated effectively. Therefore, accountants must attend to the communication process as well as the content of messages they wish to convey. Computers are increasingly being used as a communication medium for accounting information so accountants should consider the effect of human-computer interface characteristics on communication. Prior research has evaluated the effect of accounting information format, graphical or tabular, on performance of decision makers. However, a theoretical basis for performance differences has not been provided. This dissertation developed a theoretical foundation based on Dual Coding Theory to help explain why different message formats and presentation modes might differentially affect performance in tasks related to decision making. Bloom's (1956) Taxonomy of Educational Objectives was used to classify tasks by cognitive complexity.; Based on this theoretical foundation, two aspects of accounting information, message format (graphical or tabular) and presentation mode (auditory or visual), were evaluated. A computer display of financial accounting information was provided in different message formats and presentation modes to undergraduate business students. The students then answered a series of questions based on information that they had viewed and/or heard.; The results indicated little difference in the performance of subjects who had viewed graphs rather than tables. Only when the graphical message format was presented alone or supplemented with auditory information was performance significantly different from tabular message format.; While a Dual Coding effect was not supported in this research, a theoretical basis was provided that will allow replication of this experiment so that results can more easily be compared.