Forecasting the success of megaprojects with Judgmental methods

Forecasting the success of megaprojects is a very difficult and important task because ofthe complexity of such projects, as well as the large capital investment that is required forthe completion of these projects. One could argue that forecasting is not needed in thiscontext: the master Gantt chart of the tasks with assigned person-hours plus therespective Bill of Materials should suffice for an accurate estimation of the duration andcost of a project. If that was the case then every project would finish on time and on budget– but this is far from true as the numerous examples attest: HS2, Channel Tunnel, majorIT public projects in NHS, to name a few. In this research, we employ judgmentalforecasting methods to predict the success of megaprojects in as series of forecastingexperiments. In the first experiment,the participants forecast for one megaproject ('spaceexploration') with Unaided Judgment (UJ), Structured Analogies (SA) and InteractionGroups (IG) with IG showing the best results since IG>SA>SA. In the second experiment,we use a second megaproject ('a major recreational facility in the very city centre of amajor cosmopolis') and see separately the success in terms of excesses in the budget andthe duration of the project. Furthermore, the participants forecast the extent to which thesocio-economic benefits are realised. We do analyse three different stakeholderperspectives: that of the a) project manager, b) funder(s), and c) the public. We do controlfor two levels of expertise – novices, and semi-experts, and the participants use UJ, SA, IGand Delphi (D) as well, resulting IG>D>SA>UJ. In the third and final experiment, wequalitatively explore the use of scenarios in forecasting the success of megaprojects.

Level dc:type.qualificationlevel

Doctoral (Level 8)

Year dc:date.issued

2021

Author dc:creator

• Litsiou, K

Language dc:language

en

Identifier

oai:salford-repository.worktribe.com:1336307

OAI identifier oai:identifier

oai:salford-repository.worktribe.com:1336307