Content area
Full Text
Abstract ID: 140
Abstract
In practice, data envelopment analysis (DEA) often deals with estimated proportions as model inputs, which is an extension to the classic model assumptions. Several approaches recently have been proposed in the literature for each of these problems, although these often tend to be numerically intensive, Monte Carlo based, and/or require development of special purpose software to solve. We illustrate spreadsheet-based software, which solves basic conventional models as well as Monte Carlo DEA models.
Keywords
Proportions, estimated rates, Monte Carlo, data envelopment analysis, performance measurement, VBA
1. Introduction
Introduced by Charnes, Cooper, and Rhodes [1], DEA is a production frontier estimation method that solves a series of transposed fractional programs in order to determine the relative efficiency of multiple systems (e.g., hospitals) at consuming multiple inputs (e.g., doctors, resources, staff) to produce multiple outputs (e.g., clinical outcomes, access, satisfaction). DEA has been successfully used to study a variety of healthcare issues, such as hospital performance [2], public policy [3], and cardiac surgeon performance [4]. In many healthcare and other applications, DEA deals with estimated proportions. Sampling error in applications involving estimated rates or proportions can translate to significant error in the estimated efficiency scores, targets, weights and reference sets [5]. Given the importance of this problem, new methods using Monte Carlo methods have been developed [6], although the lack of an automated tool for Monte Carlo DEA has limited its applicability. We illustrate Microsoft Excel VBA based DEA software, that solve conventional and Monte Carlo DEA models.
2. DEA Solver
The program prompts for model orientation (input or output oriented), returns-to-scale relations (constant returns-to scale (CRS), variable returns-to-scale (VRS), non increasing returns-to-scale (NIRS), and non decreasing returns-toscale (NDRS)), number...