Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/15952

Title: Data Envelopment Analysis for Composite Indicators: A Multiple Layer Model
Authors: SHEN, Yongjun
WETS, Geert
Issue Date: 2013
Citation: SOCIAL INDICATORS RESEARCH, 114 (2), p. 739-756
Abstract: The development of a composite indicator (CI) over a set of individual indicators is worthwhile in case the methodological aggregation process is sound and the results are clear. It can then be used as a powerful tool for performance evaluation, benchmarking, and decision making. In this respect, data envelopment analysis (DEA), as a self appraisal technique, has recently received considerable attention in the construction of CIs for policy analysis and public communication. However, due to the ever increasing complexity of numerous performance evaluation problems, more and more potential indicators might be developed to represent an evaluation activity in a more comprehensive way. These indicators might also belong to different categories and further be linked to one another constituting a multilayer hierarchical structure. Simply treating all the indicators to be in the same layer as is the case in the basic DEA model thereby ignores the information on their hierarchical structure, and further leads up to weak discriminating power and unrealistic weight allocations. To overcome this limitation, a multiple layer DEA-based CI model is developed in this study to embody a hierarchical structure of indicators in the DEA framework, and both its primal and dual form are realized. The proposed model is illustrated by constructing a composite road safety performance index for a set of European countries.
URI: http://hdl.handle.net/1942/15952
DOI: 10.1007/s11205-012-0171-0
ISI #: 000325373100032
ISSN: 0303-8300
Category: A1
Type: Journal Contribution
Validation: ecoom, 2014
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
artikel419.95 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.