Statistical Methods in Bidding Decision Support for Construction Companies

On the border of two phases a building life cycle (LC), programming phase (conception and design) execution phase, contractor is selected. A particularly appropriate method selecting for construction market tendering system. It usually based on quality price criteria. The latter may involve (namely, direct costs connected with works realization as well mark-ups, mainly overhead profit) or cost (based costing (LCC) efficiency). contractor’s decision to participate in tender calculate requires an investment time company resources. As this often made limited frame experience subjective judgement contractor, number models have been proposed literature support process. present paper proposes use statistical classification methods. response obtained from model recommendation not. database consisting historical data was used analyses. Two were proposed: LOG model—using logit regression LDA linear discriminant analysis, which obtain better results. In model, equation function sought by indicating statistically significant variables. For purpose, backward stepwise applied, where initially all input variables introduced, namely, 15 identified bidding factors, then subsequent steps, least removed. Finally, six (factors) that significantly discriminate between groups: type works, contractual conditions, project value, need work, possible participation subcontractors, degree difficulty works. using analysis achieved good performance. results prove it can be practice. should emphasized, however, mathematical cannot replace decision-maker’s thought process, but they increase effectiveness decision.