Development of a Dispute Prediction Model for Extension of Time Dispute in the Construction Industry of Pakistan

Abstract

A construction project's performance can be quantified using important characteristics such as completion on time and on budget, if the project has been completed as per specifications, whether the project fulfilled the client's expectations, and the level of variations. Construction projects are repeatedly apprehended in a complex and indeterminate nature, and their performance may be highly susceptible to several uncertainties. This is because construction involves many construction activities, different construction companies or firms and different individuals having different knowledge and skills. This complex changing environment often gives rise to conflict, which may impact the successful completion of the project. The claims for extension of time and additional cost are one of the demands from contractor that naturally cannot always be granted by the employer, and as the result, the employer experiences difficulties in taking technical decisions, even though they have been accompanied by supervising consultant. The inherent difference of interest among parties makes a conflict inevitable which if not settled may escalate to a claim and delay or reluctance in settling claims causes disputes. Conflicts between parties must exist for a dispute to develop, and one party must make a claim while the other rejects it. Construction disputes vary in nature, size and complexity and can be costly in terms of time and money thus, the construction industry still struggles to find methods to resolve them fairly and economically. It is commonly accepted that the best solution against disputes to avoid them and necessary action to avoid dispute can be achieved only by prediction. It is well acknowledged that the global construction industry for many years has been plagued by the occurrence of delays and these delays generate negative consequences not only for project stakeholders but also for the wider construction industry and may further lead towards dispute. Delays caused by the actions of stakeholders may result in significant disputes and conflicts amongst the participants. Disputes frequency is increasing internationally, which seem to be the basis for schedule and cost overrun, with x regard to the resources utilized in their resolution. The construction industry of Pakistan is also no exception to it. To prevent and reduce disputes, the sector requires the development and implementation of appropriate decision-support systems. Thus, this study evaluated the individual causal factors of EOT or delays and developed a prediction model that may help in the early prediction of disputes in the planning phase or before it has occurred and provides a quotient/index value to predict under certain category which factor will most likely be causing dispute. The identification of factors is done through extensive review of literature and a list of factors were highlighted that were the most prevalent cause of delays. Pareto analysis was applied to the highlighted factors to find out the most contributing factors of delays in reference to literature and were shortlisted to four (variation orders, design errors, schedule delays and payment delays). With the discovery of the most contributing factors, an extensive literature review was conducted on the root cause of the shortlisted factors. The review highlighted a number of factors, but all couldn’t be studied, thus, Pareto analysis was conducted for shortlisting to most prevalent factors in each group. After the finalization of factors in each group, questionnaire was designed to check the adequacy of the factors shortlisted so far. The adequacy was checked through structured interviews with industry and academia experts. Experts gave a positive response towards the groups and factors, some factors were removed and some added based on their experience. After the finalization of the questionnaire, it was tested on selected participants, to check the representation of factors in industry and the comprehensiveness of the questionnaire. This process is called Pilot study. After the finalization of questionnaire, rating scales were selected. Due to its briefness, the 5- point Likert scale has mostly been adopted to encourage high response rate, based on the propensity of a factor towards dispute (very low – very high). To select suitable experts who have adequate knowledge and expertise to respond to this survey, two main non-probability sampling approaches, namely, purposive and snowball, were adopted. The questionnaire were xi distributed to the agreed participants and the responses were collected via mail, email, in- person and Google forms. In total (142) questionnaire were received giving a responses rate of (47.3%). The experts who participated belonged to different organization, (i.e. (26.8%) consultant, (19%) client, (38.7%) contractor and (15.5%) academic/research institutes) and (64.7%) of the respondents were involved in more than 3 high-rise building projects. The overall Cronbach’s alpha value of the four categories is 0.982 suggesting a good internal consistency reliability of the instrument, which is above the recommended threshold of To check the degree of agreement among each respondent group, Spearman’s coefficient test was conducted i.e., Spearman’s coefficient value ranges between -1 and 1, where, -1 implies perfect disagreement (strongly negative relationship), while 1 results from perfect agreement (strong positive relationship). The results depicted that almost all the values were above 0, which represents a positive relation among them. After the validation of the instrument and administering the agreement of respondents, Fuzzy Synthetic Evaluation (FSE) was employed. FSE modeling was utilized for the development of prediction model based on the survey, interviews, or pilots study results. It involved the computation of mean score, membership function of factors and groups, evaluation matrix and finally, a DPI equation for each group in each category. Variation orders, design changes and schedule delays were found to be the most critical cause of EOT or delay, further, schedule delay (people related factors) has the highest propensity towards dispute with DPI value of 3.620 followed by payment delays (process factors) with DPI 3.617.