Overview on Data Envelopment Analysis
The chapter provides a general introductory overview of data envelopment analysis. Its main purpose is to introduce the reader to the major concepts underlying this nonparametric technique. After familiarizing the reader with the general process used in c
- PDF / 416,679 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 5 Downloads / 247 Views
Overview on Data Envelopment Analysis Sergey Samoilenko
The chapter provides a general introductory overview of data envelopment analysis. Its main purpose is to introduce the reader to the major concepts underlying this nonparametric technique. After familiarizing the reader with the general process used in calculating the scores of relative efficiency, the chapter presents an overview of various orientations and types of DEA models. In conclusion, the chapter gives an overview of using DEA for the purposes of constructing Malmquist index, a popular tool for measuring changes in efficiency over time; a brief example is used to illustrate major points.
1 Introduction Data envelopment analysis (DEA) is a nonparametric method of measuring the efficiency of decision-making units (DMU). Any collection of similar entities could comprise a set of DMUs and be subjected to DEA, as long as the chosen entities transform the same type of inputs into the same type of outputs. Inputs and outputs, taken together, constitute a common DEA model for all DMUs in the sample. Thus, for all intents and purposes of DEA, every DMU in the sample is represented completely by the values of its inputs and outputs of the DEA model. Because some of the inputs or outputs of the DEA model could be more significant than others, DEA offers a decision maker a flexibility of assigning various weights to the inputs and outputs of the model; the equal weighting is commonly utilized as a default. The empirical foundation of DEA eliminates the need for some of the assumptions and limitations of traditional efficiency measurement approaches. As a result,
S. Samoilenko (*) Department of Computer Science, Averett University, 420 W Main St, Danville, VA 24541, USA e-mail: [email protected]
K.-M. Osei-Bryson and O. Ngwenyama (eds.), Advances in Research Methods for Information Systems Research, Integrated Series in Information Systems 34, DOI: 10.1007/978-1-4614-9463-8_11, © Springer Science+Business Media New York 2014
139
140
S. Samoilenko
DEA could be used in cases where the relationships between the multiple inputs and multiple outputs of the DEA model are complex or unknown. Consequently, a DEA model is not necessarily comprised of the real inputs that are converted into the real outputs as it is implied by a production process. Rather, a DEA model is better perceived as a collection of the inputs that are in some way or form important to the outputs of the transformation process under an investigation of a decision maker.
2 The General Idea Behind the Approach The original DEA model was introduced in 1978 by Charnes, Cooper, and Rhodes, and it is commonly called the CCR Model (an abbreviation consisting of first letters of the authors’ names). This model allowed representing multiple inputs and outputs of each DMU as a single abstract “meta-input” and a single “meta-output.” Consequently, the efficiency of each DMU could be represented as a ratio of the abstract input to the abstract output, and the resulting efficiency value could t
Data Loading...
