Performance evaluation of participating nations at 2012 London Summer Olympics by a two-stage data envelopment analysis

摘要

A nation's performance at Olympics can be judged by the number of medals suitably weighted by demographic factors such as population. Data envelopment analysis (DEA) is commonly used to evaluate the relative efficiency of the participating nations in the Olympics. DEA is a non-parametric technique for these measures using decision-making units (DMUs). This technique is popular since no assumptions are made on production function and no weights are imposed on inputs and outputs. There are two types of DEA-based Olympic studies made so far namely, constant input model in which each nation's input is assumed to be constant, and DEA models in which inputs vary with nations based on its social and economic conditions. Though many DEA based studies are available, they all approached the problem assuming that the nation is a black box. Instead if factors such as army recruitment, education, banking, healthcare, IT etc are also taken in to account for DMUs, then the Two-stage DEA will be more meaningful. The first stage of the two-stage model considers cultivation, training, and selection. The greater the population is, more can be the participating number of athletes. The model considers GDP per capita as most important economic element for each nation that participates. The GDP per capita and the population are the two inputs at the AP stage. The output of athletic preparation (AP) stage is the input to the second stage namely the number of participating athletes. In the athlete competition (AC) stage the above is the input and the numbers of gold, silver, and bronze medals are the outputs. This study extends the relational model (Kao & Hwang, Ref. 1) or the centralized model (Liang, Cook, & Zhu, Ref. 2) to measure the two-stage Olympic process and individual stages for each nation. (36 refs.)