Abstracts – Browse Results

Search or browse again.

Click on the titles below to expand the information about each abstract.
Viewing 1 results ...

Cheung, S O (1998) Project dispute resolution satisfaction of construction clients in Hong Kong, Unpublished PhD Thesis, School of Engineering and the Built Environment, University of Wolverhampton.

  • Type: Thesis
  • Keywords: artificial neural network; dispute resolution; Hong Kong; multivariate discriminant analysis
  • URL: http://hdl.handle.net/2436/89113
  • Abstract:

    Inspired by the concern over the dispute epidemic within the construction industry in Hong Kong, this research study sets out to examine the issue of construction dispute resolution from a management perspective. The main objectives of this research include the identification, through in-depth investigation and exploration, of the critical attributes that distinguish between Good and Bad projects in terms of their project dispute resolution satisfaction, and the development of a quantitative model for the prediction of same. Project dispute resolution satisfaction is defined as Good if the largest dispute was resolved within the site level. The novel application of the fault tree analysis provides the conceptual definition of a construction dispute. Variables selection for the model development was accomplished through a systematic and comprehensive literature review. The sustainability of the questionnaire so derived was successfully tested by a pilot study with industry experts, bringing into the research the valuable local and practical dimensions. The research problem as presented fits nicely with the classification and prediction ability of the multivariate discriminant analysis and artificial neural network multi-layer perceptron modeling. Both techniques were employed in this research for the identification of critical variables and the model development. The verification of the critical variables was carried out through a relative importance index study. The use of a principal component analysis groups the discriminating variables into three factors represented as a three level influence diagram. The relative importance index study confirmed the discriminating variables identified in the multivariate discriminant analysis model function. These variables are: The average change of tender price index; The claim consciousness of the contractor; The degree of design changes; The relationship between the project personnel; The use of alternative dispute resolution as provided in the contract; The degree of involvement of an external claim advisor by the client; The incentive of the client to settle the dispute during the resolution process; The degree of involvement of senior management during the resolution process. Prediction of project dispute resolution satisfaction by the model function achieved an accuracy of 76.92% on the hold-out sample. The sensitivity analysis of the mean for the ’Best’ network obtained from the artificial neural network multi-layer perceptron modeling suggests that the degree of design changes could well be the pivotal root to all the problems associated with construction dispute resolution.