|
Review on Parametric and Nonparametric M (2)
Review Article
1/7
Copyright © All rights are reserved by Erkie Asmare Beyene.
Volume 2 - Issue - 2
Open Access
Biostatistics & Bioinformatics
C
CRIMSON PUBLISHERS
Wings to the Research
ISSN 2578-0247
Abstract
Efficiency analysis using parametric and nonparametric methods have monopolized the recent literature of efficiency measurement. However,
the choice of estimation method has been an issue of debate. This review was aimed to: provide information on the concepts, types and methods
of applying parametric and nonparametric methods of efficiency analysis and review on the advantages and disadvantages of the two methods of
efficiency analysis. Parametric methods of efficiency analysis have significant advantages by distinguishing and modeling the random noise from
inefficiency. However, this method requires specification of the model and separating random noise from the true in efficiency may be restrictive
in most cases. On the other hand, the nonparametric method has the potential to impose axiomatic properties and estimate the frontier non-
parametrically.
In addition, it has gained popularity because of it does not have as many restrictive assumptions as parametric method. However, nonparametric
methods have deficiencies; because of it does not distinguish between the true inefficiency and statistical noise effects. Therefore, the full distance
from a brand to the efficiency frontier is interpreted as inefficiency. Generally, the parametric technique is more attractive when the data suffer
from serious measurement errors and random events. On the other way, nonparametric may be a better choice when random disturbances are
less of an issue. Therefore, a parametric and nonparametric method is not direct competitors but rather complements: in the trade off between
parametric and nonparametric, something is sacrificed for something to be gained. Hence, joint use of parametric and non-parametric techniques to
the measurement of efficiency is a novel issue in efficiency study.
Keywords:
Efficiency analysis; Stochastic frontier; Random noise; Data envelopment analysis
|
|
|