An analysis of the contractors' estimating process

Typically, contracts in the Construction Industry are placed On a basis of competitive tenders. The ability to estimate costs in advance is, therefore, imperative. Ideally, estimates should be based on recorded levels of performance but the diversity of the projects encountered often makes this feedback impractical. It has been observed that currently estimators depend to.a large extent on intuition to produce prices which are acceptable to the client. Whilst these prices are typically within 10% of the total cost of the work, the estimates for individual work items may differ from the actual cost by as much as 100%. For this reason, research was undertaken into methods which are not normally practiced in the Construction Industry; The principal objective of the research was to find and test a method of estimating which would be quicker than present methods and would achieve a more direct reflection of observed performance. This thesis describes attempts at applying regression analysis globally to building costs. It also examines the varying levels of difficulty in estimating different categories of work and the implications of the Pareto distribution of the values of the items. in a bill of quantities. [Continues.]