IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 11, Issue 3 Ver. VI (May- Jun. 2014), PP 43-51 www.iosrjournals.org www.iosrjournals.org 43 | Page “Suggested Mathematical Model for Specialized Subcontractor Prequalification Scrutiny and Ultimately the Performance Prediction” Mr. Mangesh M. Kapote 1 , Prof. Dr. S. S. Pimplikar 2 1 Student, M.E. (Construction & Management), Maharashtra Institute of Technology, Pune, India. 2 Professor & Head, Civil Engg. Dept ,Maharashtra Institute of Technology, Pune, India. Abstract: A large portion of the work done in construction projects is carried out by subcontractors. Thus it is very much essential to select a right subcontractor for project’s success. Based on dependent, independent variables and their correlation, Weighted Point Score Method for prequalification scrutiny of subcontractor is suggested here with a Logistic Regression (LR) approach which ultimately helps in performance prediction of specialized subcontractor. The paper summarizes the Logistic Regression approach and its advantages, also the Weighted Point Score Method with quantification of different variables on 3 point scale which helps in selecting best suited subcontractor. Keywords: Dependent Variable, Independent Variables, Logistic Regression (LR) Approach, Prequalification, Specialized Subcontractor, Weighted Point Score Method. I. Introduction Subcontracting is advantageous for the general contractors in many ways such as reduced capital investment for general contractors, speedy work, good quality work, reduced risk to general contractors, etc. Although the success of a project does not depend wholly on judicious subcontractor selection, choosing the right subcontractor is important because many defaults in the past have been due to subcontractors accepting jobs, they are incapable of undertaking and good subcontractors being given inappropriate contracts (Okoroh and Torrence 1999; Kumarswamy and Matthews 2000). Several sophisticated methods have already been proposed for the selection of main contractors and subcontractors such as multicriteria decision making, multiattribute analysis, multiple regression, cluster analysis, fuzzy set theory, multivariate discriminant analysis (Hatush and Skitmore 1997; Albino and Garvelli1998; Holt 1998; Mahdi et al. 2002). New findings pertaining to contractor prequalification, tender evaluation and modeling techniques for predicting contractor‟s performance are confirming that the subject area still justifies the investigation (eg., Abidali and Harris 1995; Tam and Harris 1996; Chinyio et al. 1998; Ng et al. 1999; Lam et al. 2000; Wong and Holt 2001). This paper presents a weighted point score method for prequalification scrutiny of subcontractors with logistic regression (LR) approach. The dependent and independent variables are identified from literature survey and weightages are given to the independent variables by discussing with field professionals and literature survey analysis. Formulation of Mathematical Model: For convenience the independent variables are grouped in 8 catagories so as to generate the final form of the logistic regression as given below, Y = Dependent Variable = The selection of an interested bidder to work as a subcontractor with a reputed contracting firm adopting a process based approach depends upon. X = Independent Variable X1 = Staff quality and Experience X2 = Plant and Equipment resources X3 = Subcontractor site management / execution capability X4 = Health and Safety X5 = Past performance records on similar projects X6 = Subcontractor Reputation / Image X7 = Subcontractor Proposal X8 = Other Evaluation criteria The Logistic Regression (LR) technique is used to determine how the probability of a subcontractor performance can be predicted (good or poor) from their previous completed projects. In arithmetic terms, this relationship takes the form of,