Human Development Index (HDI) Rank-Order
Variability
Ahmet M. C¸ ilingirtu
¨
rk
1
•
Habip Koc¸ak
2
Accepted: 9 March 2017 / Published online: 29 March 2017
Ó Springer Science+Business Media Dordrecht 2017
Abstract In accordance with the increasing demand for information, indices are created
and national and global rankings made to represent and through which to understand and
build policy related to complex situations, processes and trajectories. Different indices for
a single concept are also created that have advantages or disadvantages over one another or
to overcome certain calculation problems. As one such, the Human Development Index
(HDI) presently lists countries according to four different criteria, and remains at the heart
of democratic and humanitarian recovery efforts. This type of indicator is taken as a
function of past performances, with high performances being the extreme values at posi-
tively skewed distributions. Thus, the variability of each unit’s repeated measures is
regarded as the result of efforts made between the measurement time points (in the HDI
case, of a country to promote development). However, it is assumed that the variability of
the units is not homogenous. Here, it is shown that in the HDI case, high performance units
show relatively low variability, whereas the middle and middle-low performance units
show a high variability. Cluster analysis and Friedman test have been used to determine the
characteristics of ordered country rankings. The variability of rank-order should also be
taken into account besides the location on the list by clustering the countries according to
HDI.
Keywords Human Development Index Á HDI rank Á Order variability Á Superstars theory Á
Friedman Chi square Á Clustering
& Ahmet M. C¸ ilingirtu
¨
rk
acilingi@marmara.edu.tr
1
Department of Econometrics, Statistics, Marmara University, Goztepe Campus, Kadıkoy,
34722 Istanbul, Turkey
2
Department of Econometrics, Operations Research, Marmara University, Istanbul, Turkey
123
Soc Indic Res (2018) 137:481–504
https://doi.org/10.1007/s11205-017-1605-5
@Phil_Robichaud
@deepthiw
@JoseServera