A Model for Characterizing Tender Behavior

• Activity Number: 246
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Sollie Millard*+ and Frans H.J. Kanfer and Andriette Bekker and Mohammad Arashi
• Companies: University of Pretoria and University of Pretoria and University of Pretoria and Shahrood University
• Keywords: Bayesian inference ; finite mixture model ; tender behaviour

Abstract

Government and other entities typically source external services on a tender basis. For a long time it was, and in many cases still is, standard practice to appoint the service provider based on purely the lowest price. Such an approach has several disadvantages. Tender behaviour is modelled by fitting appropriate distributions to the historical data, and using these models to select an appropriate service provider based on pricing that corresponds to a pre-specified percentile. In this presentation a finite mixture model is proposed for modelling the tender data, considering skewed distributions for the mixing components. Is this approach a promising alternative for the characterization of tender behaviour?