Use this URL to cite or link to this record in EThOS:
Title: Homogeneity in data envelopment analysis
Author: Wyatt, Ross
Awarding Body: University of Kent
Current Institution: University of Kent
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
Data Envelopment Analysis (DEA) is a non-parametric approach to measuring efficiency. It is increasingly utilised in areas such as finance, education and healthcare. As with many modelling techniques, DEA is based on and restricted by certain assumptions. One of these is the assumption of homogeneity, which asserts that all decision making units must have the same inputs and outputs. Previously, research has had to develop ad hoc methods to cope with this issue, such as removing non-homogenous units, removing key inputs and outputs, having to run multiple DEA models, and using other methods such as regression analyses and post-hoc comparisons. The assumption of homogeneity has therefore restricted the utility of DEA as a measure of efficiency and has limited the scope of its application in the research field. This thesis presents the development and application of a new method that overcomes this assumption, in the form of the 'non-homogeneity' model. The model is an extension of the multi-activity model (Beasley 1995, Mar Molinero 1996) and allows units to be compared when they do not share the same inputs and outputs. The model is firstly applied to an existing dataset to demonstrate its properties. Secondly, to test its usefulness in different scenarios, the model is applied using the concept of indexes, to determine whether new insights into the analysis of indexes can be identified. Finally, the model is used to analyse the efficiency of microcredit providers in the Bolivian financial sector. This complex, non-homogeneous data provides an ideal context to test the non-homogeneity model. It also demonstrates the utility of the model in being able to identify whether units are better to specialise in an area or diversify. This research furthers the field of microcredits by incorporating many different types of lenders into a single analysis. This thesis provides a significant contribution to the DEA literature as it firstly presents an opportunity to analyse non-homogeneous units, and secondly allows researchers to identify whether it is better for those units to specialise or diversify.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available