Graphical method for identifying high outliers in construction contract auctions

Abstract

Construction contract auctions are characterised by (1) a heavy emphasis on the lowest bid, as that is which usually determines the winner of the auction, (2) anticipated high outliers due to the presence of uncompetitive bids, (3) very small samples, and (4) uncertainty of the appropriate underlying density function model of the bids. This paper describes a graphical method for simultaneously identifying outliers and density functions by first removing candidate (high) outliers and then examining the goodness-of-fit of the resulting reduced samples by comparing the reduced sample predictability (by the expected value of the lowest order statistic) of the lowest bid with that of the equivalent predictability by Monte Carlo simulations of one of the common density functions. When applied to a set of 1073 auctions, the results indicate the appropriateness of censored and reduced sample lognormal models for a wide range of cut-off values. These are compared with cut-off values used in practice and to identify potential improvements.