information on the factors identified in the literature as the most influential:
the socio-economic and family environment, the innate ability of the
students and the characteristics of their class mates (peer group effect).
With respect to the technique used to measure the relative
efficiency of schools, two main alternatives can be consid
ered: parametric
and non-parametric methods
. In the literature
, on the one hand,
some studies comparing efficiency scores gen
erated by both techniques for a
specific sample (Bates, 1997;
Chakraborty et. al, 2001; Mizala et. al, 2002)
may be found. On the other, there are studies
using Monte Carlo experiments where the underlying production technology is known (Yu,
1998). Nevertheless, most authors use non -parametric approximations
and, specifically, Data Envelopment Analysis (DEA)
. This choice is based, a mongst other reasons, on its great flexibility, which makes it
particularly suitable in an area such as education where the production
function is unknown, and on its ability to adapt to processes involving not
only a range of inputs but also a series of i
ntermediate outputs, rather than
a single final input
. Moreover, in recent years different methods have
been developed to incorporate in the technique the fact that there are non
controllable inputs when efficiency scores are calculated, which is of
particular interest in the educational sector.
DEA, introduced by Charnes, Cooper and Rhodes (1978), is
characterised by the fact that it does not impose a specific functional form
on the production function, but rather establishes certain assumptions
about the properties of technology which allow the definition of the set of
feasible productive processes whose frontier envelops the observed data.
The standard formulation of the programme can take several forms
according to different criteria, so it can be or
iented to reduce input values