An experimental investigation of the effects of artificial intelligence systems on the training of novice auditors

Wongpinunwatana, Nitaya. (1998). An experimental investigation of the effects of artificial intelligence systems on the training of novice auditors PhD Thesis, School of Business, The University of Queensland.

       
Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
THE12712.pdf Full text Click to show the corresponding preview/stream application/pdf 14.19MB 2
Author Wongpinunwatana, Nitaya.
Thesis Title An experimental investigation of the effects of artificial intelligence systems on the training of novice auditors
School, Centre or Institute School of Business
Institution The University of Queensland
Publication date 1998
Thesis type PhD Thesis
Supervisor Associate Professor Colin Ferguson
Dr Paul Bowen
Total pages 255
Subjects 15 Commerce, Management, Tourism and Services
Formatted abstract The primary objective of this thesis is to investigate the impact of task-technology fit on novices' performance when using artificial intelligence systems for auditing tasks. Understanding the effect of task-technology fit on novices is important because appropriate task-technology fit can assist organisations in maximising the benefits of developing intelligent tutoring systems.

This research develops a model of the impact of task-technology fit on novices. The model is based on three theories: the memory model, the mental models, and the cognitive fit model. The task-technology fit model predicts that novices employing a high task-technology- fit system will solve problems more accurately and require less time to solve problems than novices employing a low task-technology-fit system. In particular, the matching of an appropriate artificial intelligence system to individual tasks is hypothesised to influence novices' certainty of the correctness of their solutions, confidence in the conclusions provided by the artificial intelligence systems, and perceptions of the quality of the artificial intelligence systems.

Four artificial intelligence auditing systems, two problem-solving programs, and four questionnaires were developed. A laboratory experiment was performed with two hundred and ninety-two undergraduate auditing students. The students used rule-based reasoning and case-based reasoning systems to learn how to make internal control over purchases evaluations (a structured task) and going concern decisions (an unstructured task). The students then completed six-problem solving cases without using the artificial intelligence systems. The answers from problem-solving cases and questionnaires were analysed.

The results suggested that the effect of task-technology fit on accuracy in solving problems was marginal for case-based reasoning with unstructured tasks. No significant effect was found on problem-solving accuracy for rule-based reasoning with structured tasks. Task-technology fit did not affect length of time in solving problems. The task-technology fit, however, marginally increased novices' certainty of the correctness of their solutions. The task-technology fit had an effect on confidence in conclusions provided by the artificial intelligence systems and perceptions of ease of use of the artificial intelligence systems only for rule-based reasoning with structured task. For case-based reasoning with unstructured tasks, the opposite direction to that expected was found on confidence in conclusions provided by the systems and perceptions of ease of use of the systems. Perceptions of usefulness of the artificial intelligence systems were not affected by task-technology fit. The results also indicated that the effect of structured tasks on accuracy in solving problems was stronger than for unstructured tasks. The analyses failed to find statistically significant results for the effect of task type on novices' length of time in solving problems, certainty of the correctness of their solutions, confidence in conclusions from the artificial intelligence systems, and perceptions of the quality of the artificial intelligence systems.

In summary, the research presented in this thesis contributes to the development of the theory of the impact of task-technology fit on novices' performance in problem solving and perceptions when using artificial intelligence systems for training in auditing domain. This study also identifies several areas for future research. Research could be conducted, for example, on comparing the effects of retrieval-only and fully automated case-based reasoning systems on novices' problem solving in the auditing domain. The impact of input dialogue structures and system explanations from case-based reasoning systems on novices' problem-solving performance could also be investigated.
Keyword Artificial intelligence.
Auditing -- Data processing.

 
Citation counts: Google Scholar Search Google Scholar
Access Statistics: 79 Abstract Views, 2 File Downloads  -  Detailed Statistics
Created: Mon, 29 Nov 2010, 16:26:08 EST by Muhammad Noman Ali on behalf of Library - Information Access Service