di Paolo Radaelli, Michele M. Zenga, Michele Zenga
di Mohd T. Alodat, D.A. Omari
In the ranked set sampling with set size equal to m, there are mm different possible sampling
schemes. The aim of this paper is to characterize the optimal sampling scheme via the Fisher information
numbers for normal distribution with known coefficient of variation. Moreover, we use the
optimal sampling schemes to obtain unbiased estimators for the mean. Also we compare the estimators
to their counterparts under the McIntyre’s ranked set sampling scheme. We show that the MLE
under optimal sampling scheme is asymptotically more efficient than its counterpart under McIntyre’s
Keywords: Ranked Set Sampling, Fisher Information, Maximum Likelihood, Normal Distribution,
Coefficient of Variation.
di K.R. Muraleedharan Nair, N. Sreelakshmi, N. Unnikrishnan Nair
In the present communication, we discuss some properties of the new Zenga measure I (p) for inequality.
Unlike the Lorenz curve, the I (p) curve can be decreasing, increasing and non-monotonic.
Various conditions for monotonicity of I (p) are presented. Further some results on a stochastic order
based on I (p) are also established.
Keywords: Zenga Curve, Lorenz Curve, Stochastic Order, Quantiles.
di Sergio Scippacercola
Today the Web Pages with static contents (e.g. on line newspapers, public portals, didactic pages,
etc.) are very widespread by Information Technology. These pages are carriers of knowledge so it
is important to study their use. The common web analytics tools lack useful metrics for an evaluation
of the dissemination of knowledge via the web. In this paper we propose a Method (called
Knowledge Page Markovian Algorithm) that uses the Markov Chains on the data obtained from
web analytics tools to evaluate the use of the knowledge pages. By this original approach, it is possible
to create new Metrics to assess each web page for the eventual reengineering of the site. Finally,
an application of the Method is shown on a website for three University Courses.
Keywords: Web Analytics, Metrics, Knowledge Page, Markov Chains.
di Camilla Ferretti
Mobility indices are important tools for summarizing and comparing the dynamics of different samples.
A mobility index is defined as a function which assigns a real value I to a given transition matrix
. The function I ðPÞ, where P is the transition matrix, is chosen to describe a specific type of
mobility (e.g. the tendency to leave the current state). In this paper we develop a family of mobility
indices capable of embodying information on both the intensity and the prevailing direction of mobility
in the sample under study.
Keywords: Mobility Index, Transition Matrix, Prevailing Direction.
di Enrico Ciavolino, Anna Maria Frasca, Piergiorgio Mossi
For several years it has become the practice of Public Administration quality assessment as part of
the institutional activities carried out by offices to improve public services by monitoring the production
process. The Ministry of Labour and Social Policies has adopted a system of ‘‘quality assessment’’
based on indicators selected as representative of the inspection carried out in decentralized
offices. This scoring system – called Project Quality – is defined by three ‘‘synthetic indicators’’ determined
periodically by the 92 Provincial Labour Directorates (in Italian: Direzioni Provinciali del
Lavoro – DPL) operating in the country. It does indeed have a rating system that defines a ranking
among the offices. This paper presents the results of the research in order to analytically describe
the performance level with a different model and also suggesting the possible influence exerted by
‘‘local context variables’’, i.e., those relating to the geo-socio-economic differentials, in explaining
the efficiency of inspection. The data are analyzed according to the variable inspection and local
was formalized through a second-order structural equations model.
Keywords: PLS Path-Modeling, Second Order Latent Variables, Performance Evaluation, Context
Variables; Inspection Activity.