Showing papers in "Statistica in 2013"
TL;DR: In this article, the authors compare seasonal minimum temperatures from model output and observations using marginal extreme value modeling techniques, and empirically estimate the shift in the distribution as a function of the regional climate model values, using the Doksum shift function.
Abstract: Using output from a regional Swedish climate model and observations from the Swedish synoptic observational network, we compare seasonal minimum temperatures from model output and observations using marginal extreme value modeling techniques. We make seasonal comparisons using generalized extreme value models and empirically estimate the shift in the distribution as a function of the regional climate model values, using the Doksum shift function. Spatial and temporal comparisons over south central Sweden are made by building hierarchical Bayesian generalized extreme value models for the observed minima and regional climate model output. Generally speaking the regional model is surprisingly well calibrated for minimum temperatures. We do detect a problem in the regional model to produce minimum temperatures close to 0◦C. The seasonal spatial effects are quite similar between data and regional model. The observations indicate relatively strong warming, especially in the northern region. This signal is present in the regional model, but is not as strong.
15 citations
TL;DR: In this article, the authors used generalized additive models to evaluate the spatio-temporal variation of both species, together with the possible nonlinear effects of biotic and abiotic factors.
Abstract: In theMediterrean Sea the population features of demersal resources fluctuate over spatial and temporal scales due to the variability of abiotic and biotic factors as well as to human activities. The two shrimps Parapenaeus longirostris and Aristaeomorpha foliacea are among the most important deep-sea demersal resources in the North-Western Ionian Sea. Their changes in terms of density, biomass andmedian length induced by anthropogenic and environmental variables (fishing effort, sea surface temperature, precipitations, Winter North Atlantic Oscillation (NAO) and Annual MediterraneanOscillation (MO) indices) were investigated. Biological data were collected during trawl surveys carried out from 1995 to 2006 as part of the international program MEDITS (International Bottom Trawl Survey in the Mediterranean). Generalized AdditiveModels were used to evaluate the spatio-temporal variation of both species, together with the possible nonlinear effects of biotic and abiotic factors. Density and biomass were assumed to be distributed according to a member of the Tweedie family in order to account for zero-inflation in the relative data. Spacetime interaction was consideredwithin a non-separablemodel with smooth spatio-temporal component based on tensor product splines. The results show significant spatio-temporal and depth effects in the three population parameters of these resources. Winter NAO index significantly influenced the density, biomass and length of P. longirostris. Sea surface temperature significantly influenced the size of this species and the three population features of A. foliacea. The size of this shrimp resulted also influenced negatively by fishing effort and positively by the MO index.
14 citations
TL;DR: In this article, the authors proposed two mixed randomized response techniques as an alternative to the Kim and Warde's (2005) randomized response technique, and the properties of the models have been studied.
Abstract: The paper proposes two mixed randomized response techniques as an alternative to the Kim and Warde’s (2005) randomized response technique. The properties of the models have been studied and found that the proposed mixed randomized response models are better than the Kim and Warde’s (2005) mixed randomized response models in some realistic situations. We extend the proposed model to stratified sampling. Numerical illustration is given in support of the present study.
12 citations
TL;DR: In this paper, the authors study the spatio-temporal point patterns of wildfires and model them by a log Gaussian Cox process, which is used in the narrow sense as the annual fire risk map for next year.
Abstract: Policy responses for local and global firemanagement depend heavily on the proper understanding of the fire extent as well as its spatio-temporal variation across any given study area. Annual fire risk maps are important tools for such policy responses, supporting strategic decisions such as location-allocation of equipment and human resources. Here, we define risk of fire in the narrow sense as the probability of its occurrence without addressing the loss component. In this paper, we study the spatio-temporal point patterns of wildfires and model them by a log Gaussian Cox processes. Themean of predictive distribution of randomintensity function is used in the narrow sense, as the annual fire risk map for next year.
12 citations
TL;DR: In this article, a modified form of ZILSD was developed and several important aspects of it were derived and some test procedures were suggested, and the parameters of modified ZILDS were estimated by method of maximum likelihood and illustrated the procedure using certain real life data sets.
Abstract: Here we consider a zero-inflated logarithmic series distribution (ZILSD) and study some of its properties. A modified form of ZILSD is also developed and derived several important aspects of it. The parameters of modified ZILSD are estimated by method of maximum likelihood and illustrated the procedure using certain real life data sets. Further, certain test procedures are suggested.
11 citations
TL;DR: In this paper, the authors provide a general picture of the main biodiversity indices proposed in literature and show their limits in favor of diversity profiles, and point out the advantages of the parametric FDA which, under suitable assumptions, allow us to obtain summary statistics of the same functional form of the observed data.
Abstract: The past few years have highlighted the need to protect diversity in its broader concept and, in particular, with reference to the ecological context for environmental protection. In this context, the problem of the definition and the measurement of diversity becomes fundamental. This paper provides a general picture of the main biodiversity indices proposed in literature and shows their limits in favor of diversity profiles. Since the diversity profile is a curve which expresses the diversity as a function of the relative abundance vector, it may be studied through the functional approach. In particular,we point out the advantages of the parametric FDA which, under suitable assumptions, allow us to obtain summary statistics of the same functional form of the observed data. Diversity profiles and parametric FDA approach have been applied to evaluate the diversity of epiphytic lichens in the province of Florence, providing an alternative way of understanding the biological diversity.
10 citations
TL;DR: In this paper, an improved estimator of parameters of the Morgenstern type bivariate logistic distribution (MTBLD) using ranked set sampling has been proposed, which has shown the superiority of the proposed estimators over Chacko and Thomas (2009) estimators.
Abstract: In this paper we have suggested some improved estimator of parameters of Morgenstern type bivariate logistic distribution (MTBLD) using ranked set sampling. We have shown the superiority of the proposed estimators over Chacko and Thomas (2009) estimators.
6 citations
TL;DR: In this paper, the authors introduced a new class of elliptically contoured processes and found the link between the conditional time varying behavior of the covariance matrix of the returns and the time variability of the investor's coefficient of risk aversion.
Abstract: In this paper, we introduce a new class of elliptically contoured processes. The suggested process possesses both the generality of the conditional heteroscedastic autoregressive process and the elliptical symmetry of the elliptically contoured distributions. In the empirical study we find the link between the conditional time varying behavior of the covariance matrix of the returns and the time variability of the investor’s coefficient of risk aversion. Moreover, it is shown that the non-diagonal elements of the dispersion matrix are slowly varying in time.
5 citations
TL;DR: In this paper, the authors proposed to extend the LSM property in terms of probability density function and examine therefrom its characterization and stability properties, which can be interlinked with reciprocal coordinate subtangent of the density curve and a few other derived measures.
Abstract: The celebrated lack of memory property is a unique property of the exponential distribution in the continuous domain. It is expressed in terms of equality of residual survival function with the survival function of the original distribution. We propose to extend this lack of memory property in terms of probability density function and examine therefrom its characterization and stability properties. In this process the density version of the lack of memory property can be interlinked with reciprocal coordinate subtangent of the density curve and a few other derived measures.
3 citations
TL;DR: In this article, the authors discuss the conditions under which some socioeconomic variables measured on European countries tend to converge over time towards a common level towards the use of renewable sources and consider the extent to which EU countries meet the binding commitment to reach a fifth of energy consumption from renewable sources by 2020.
Abstract: In recent years there has been an increasing awareness on problems related to the economic growth and on the conditions under which some socio-economic variables measured on European countries tend to converge over time towards a common level. This paper is concerned with the use of energy from renewable sources and considers the extent to which EU countries meet the binding commitment to reach a fifth of energy consumption from renewable sources by 2020. By discussing empirical results on the economic growth pattern of 28 countries in the period 1995-2010, we make use of several spatial growth regression models. We show that the proposed models are able to capture the complexity of the phenomenon including the possibility of estimating sitespecific convergence parameters and the identification of convergence clubs.
3 citations
TL;DR: In this paper, a two-phase chain-type ratio estimator was proposed for estimating the population mean of the study variable y using the information on two auxiliary variables x and z. The bias and mean squared error were obtained upto the first order of approximation.
Abstract: This paper advocates the problem of estimating the population mean of the study variable y using the information on two auxiliary variables x and z. We have suggested the family of chain ratio exponential type estimators in two-phase (or double) sampling. The bias and mean squared error (MSE) are obtained upto the first order of approximation. The suggested class of estimators is more efficient than the two-phase ratio estimator, Chand’s (1975) chain-type ratio estimator and Singh and Vishwakarma’s (2007) estimator in two-phase (or double) sampling. An empirical study is given to justify the superiority of the proposed estimator.
TL;DR: A unified methodology for sequential clinical trials under a Bayesian paradigm is proposed and a prevision based approach to Gaussian data is applied to make predictive inference based on the data accrued so far together with future data.
Abstract: The paper deals with the Bayesian sequential analysis of clinical trials and the related predictive approach in a particularly innovative fashion. We propose a unified methodology for sequential clinical trials under a Bayesian paradigm. The idea is to make predictive inference based on the data accrued so far together with future data. We apply a prevision based approach to Gaussian data and end up with some numerical illustrations.
TL;DR: The aim of the present paper is to measure the mode effect on the answers using a new data driven multivariate approach, that allows to disentangle the interview Mode effect on answers from the effect of self selection.
Abstract: Many studies are showing an increased tendency to use more than one data collection mode for a particular survey. However, mixed data collection modes may influence responses given by interviewees and require researchers to verify if differences in responses, when present, are ascribable to the type of data collection mode. Often, random assignment is not feasible and requires researchers to solve an additional and not negligible problem, namely to verify if differences in responses are ascribable to the self selection or to the type of data collection mode being used. The aim of the present paper is to measure the mode effect on the answers using a new data driven multivariate approach, that allows to disentangle the interview mode effect on answers from the effect of self selection. We will work through the use of the new multivariate method with AlmaLaurea case concerning the evaluation of two different data collection methods: the CAWI (Computer Assisted Web Interviewing) and the CATI (Computer Assisted Telephone Interviewing). As with any new statistical method, the success of this method depends on its efficacy in relation to that of the existing methods. Therefore, results of the multivariate approach will be compared to the Propensity Score method that AlmaLaurea usually applies to identify the presence of an interview mode effect. Both methods produce similar results.
TL;DR: In this paper, a three phase sampling is proposed to estimate the population mean of character by a three-phase stratified ratio estimator and some numerical results are presented to illustrate efficiency of the proposed procedure against possible alternative ones and from the expected cost optimum sample sizes on first, second and third phases.
Abstract: In this paper a three phase sampling is proposed to estimate the population mean of character by a three phase stratified ratio estimator and some numerical results are presented to illustrate efficiency of the proposed procedure against possible alternative ones and from the expected cost optimum sample sizes on first, second and third phases are also calculated.
TL;DR: In this article, a dynamic panel analysis of the renewable investment in a sample of 29 countries is proposed to analyze the driving of investment in renewable energy sources in countries with, or without, nuclear power plants.
Abstract: The aimof this study is to analyze the driving of investment in renewable energy sources in countries with, or without, nuclear power plants. To address these issues, a dynamic panel analysis of the renewable investment in a sample of 29 countries is proposed. Results demonstrate that investments in renewable sources present an inverse relationship with share of nuclear power generation in countries with nuclear power plants but in the countries with power generation based on fossil fuel, investments in renewable sources depend on GDP and technological efficiency. Results also show that energy sustainability passes through the use of renewable resources that can complement the nuclear technology on condition that both exceed their limits.
TL;DR: In this paper, a suitable design-based approximation to the BIC is proposed in order to avoid the derivation of the exact likelihood of the sample which is often very complex in a finite population sampling.
Abstract: In this article, various issues related to the implementation of the usual Bayesian Information Criterion (BIC) are critically examined in the context of modelling a finite population. A suitable design-based approximation to the BIC is proposed in order to avoid the derivation of the exact likelihood of the sample which is often very complex in a finite population sampling. The approximation is justified using a theoretical argument and a Monte Carlo simulation study.
TL;DR: In this article, the authors focused their attention on a particular composite index of gender equality, the Global Gender Gap Index (GGGI), highlighting problematics and weaknesses and proposing a different approach structured in four steps.
Abstract: In this paper, we focus our attention on a particular composite index of gender equality, the Global Gender Gap Index (GGGI), highlighting problematics and weaknesses and proposing a different approach structured in four steps. The starting point of our analysis is to narrow the research to a small group of OECD countries: in this way it is possible to lower the gender analysis in a homogeneous socio-cultural framework and introduce a fifth dimension related to the time use. Next, to explore which variables have a greater impact on the gender gap persistence among these countries, we propose a different weighting method, based on the structural equation modeling (SEM). Through the study of official data, the effects of these steps on the final ranking of countries were then analyzed, allowing reflections from both the methodological and socio-cultural point of view.
TL;DR: Described of the official statistical information sources regarding the environment with a focus on soil and water, ISTAT and ISPRA, describing their internal structures and links to corresponding EU institutions.
Abstract: In this paper, we provide descriptions of the official statistical information sources regarding the environment with a focus on soil and water. The discussion focuses on two major contributors of official statistical information in the field, ISTAT and ISPRA, describing their internal structures and links to corresponding EU institutions. We further review their latest production of statistical information on soil and water and provide several suggestions for the potential development of the field.
TL;DR: In this paper, text mining models were used to categorize and measure the suggestions of Italian PhD holders on improving their own programs in the STELLA survey (Statistiche in TEma di Laureati e LAvoro).
Abstract: We present Text Mining models to thematically categorise and measure the suggestions of PhD holders on improving PhD programmes in the STELLA survey (Statistiche in TEma di Laureati e LAvoro). The coded responses questionnaire, designed to evaluate the employment opportunities of students and assess their learning experience, included open-ended questions on how to improve PhD programmes. The Corpus analysed was taken from the data of Italian PhD holders between 2005 and 2009 in eight universities (Bergamo, Brescia, Milano Statale, Milano Bicocca, Pisa, Scuola Superiore Sant’Anna, Palermo and Pavia). The usual methodological approach to text analysis allowed us to categorize open-ended proposals of PhD courses improvements in 8 Italian Universities.
TL;DR: In this paper, the authors describe some of the evolution in environmental regulation, management and monitoring and the information needs, closely aligned to the statistical challenges to deliver the evidence base for change and effect.
Abstract: This short article describes some of the evolution in environmental regulation, management and monitoring and the information needs, closely aligned to the statistical challenges to deliver the evidence base for change and effect.
TL;DR: In this article, the authors obtained recurrence relations for higher moments of order statistics from doubly truncated Burr distribution, which enable one to obtain all the single, double (product) and higher moment of any order of all order statistics for any sample size.
Abstract: In this paper, we have obtained recurrence relations for higher moments of order statistics from doubly truncated Burr distribution, which enable one to obtain all the single, double (product) and higher moments of any order of all order statistics for any sample size from doubly truncated Burr distribution in a simple recursive manner, thus generalizing the earlier work done by Khan and Khan (1987) and also by Pushkarna, Saran and Tiwari (2012).
TL;DR: In this article, a bivariate form of the discrete Linnik distribution is introduced and its distributional properties are obtained using compounding schemes, where autoregressive processes are developed with marginals.
Abstract: Christoph and Schreiber (1998a) studied the discrete analogue of positive Linnik distribution and obtained its characterizations using survival function. In this paper, we introduce a bivariate form of the discrete Linnik distribution and study its distributional properties. Characterizations of the bivariate distribution are obtained using compounding schemes. Autoregressive processes are developed with marginals follow the bivariate discrete Linnik distribution.
TL;DR: In this paper, an alternative methodology based on a space-time point process where the intensity function indicates the limiting expected rate of occurrence of snow avalanches occurring on day t at location (x, y), conditioned on the historical information available prior to time t.
Abstract: The study of avalanche events is particularly important to assess and predict the degree of risk involved in a given area and time. In this work we consider an alternative methodology based on a space-time point process where the intensity function indicates the limiting expected rate of occurrence of snow avalanches occurring on day t at location (x, y), conditioned on the historical information available prior to time t . The model depends also on some environmental variables (degree of slope, exposure, altitude, etc.) which may be considered as covariates. In order to show the spatio temporalmodeling of the avalanche hazardwe consider the application to the digitalized Avalanche Database of the Trentino region, Italy.
TL;DR: In this paper, a new version of FI LODE based on decomposition of a new structure of the variance-covariance matrix using singular value decomposition (SVD) instead of Spectral Decomposition (SD).
Abstract: Least Orthogonal Distance Estimator (LODE) of Simultaneous Equation Models’ structural parameters is based on minimizing the orthogonal distance between Reduced Form (RF) and the Structural Form (SF) parameters. In this work we propose a new version – with respect to Pieraccini and Naccarato (2008) – of Full Information (FI) LODE based on decomposition of a new structure of the variance-covariance matrix using Singular Value Decomposition (SVD) instead of Spectral Decomposition (SD). In this context Total Least Square is applied. A simulation experiment to compare the performances of the new version of FI LODE with respect to Three Stage Least Square (3SLS) and Full Information Maximum Likelihood (FIML) is presented. Finally a comparison between the FI LODE new and old version together with few words of conclusion conclude the paper.
TL;DR: In this paper, the authors showed that Abdelfatah et al. this paper proposed a Randomized Response (RR) technique with two decks of cards proposed by Odumade and Singh (2009) can always be made more efficient.
Abstract: The Randomized response (RR) technique with two decks of cards proposed by Odumade and Singh (2009) can always be made more efficient than the RR techniques proposed by Warner (1965), Mangat and Singh (1990), and Mangat (1994) by adjusting the proportion of cards in the decks. Abdelfatah et al. (2011) modified Odumade and Singh (2009) RR technique and claimed that their method can be more efficient than the Warner (1965) model. In this paper it is shown that such claim is not valid and the RR technique proposed by Abdelfatah et al. (2011) is in fact less efficient than the Warner (1965) technique at equal protection of respondents. Such finding are recently shown by Giordano and Perri (2011).
TL;DR: In this article, a simulation study of the sample path of the arrival process of the DAR(1)/D/s Queue with Discrete Mittag-Leffler [DML( α )] as marginal distribution is conducted.
Abstract: In this paper we analyze DAR(1)/D/s Queue with Discrete Mittag-Leffler [DML( α )] as marginal distribution. Simulation study of the sample path of the arrival process is conducted. For this queueing system, the stationary distribution of the system size and the waiting time distribution of an arbitrary packet is obtained with the help of matrix analytic methods and Markov regenerative theory. The quantitative effect of the stationary distribution on system size, waiting time and the autocorrelation function as well as the parameters of the input traffic is illustrated empirically. The model is applied to a real data on the passenger arrivals at a subway bus terminal in Santiago de Chile and is established that the model well suits this data.
TL;DR: In this article, a sampling survey framework and a model-based approach are proposed to estimate model variances through historical data concerning past survey occasions. But the model parameters and functions are not considered.
Abstract: In this context, supposing a sampling survey framework and a model-based approach, the attention has been focused on the main features of the optimal prediction strategy for a population mean, which implies knowledge of some model parameters and functions, normally unknown. In particular, a wrong specification of the model individual variances may lead to a serious loss of efficiency of estimates. For this reason, we have proposed some techniques for the estimation of model variances, which instead of being put equal to given a priori functions, can be estimated through historical data concerning past survey occasions. A time series of past observations is almost always available, especially in a longitudinal survey context. Usefulness of the technique proposed has been tested through an empirical attempt, concerning the quarterly wholesale trade survey carried out by ISTAT (Italian National Statistical Institute) in the period 2005-2010. In this framework, the problem consists in minimising magnitude of revisions , given by the differences between preliminary estimates (based on the sub-sample of quick respondents) and final estimates (which take into account late respondents as well). Main results show that model variances estimation through historical data lead to efficiency gains which cannot be neglected. This outcome was confirmed by a further exercise, based on 1000 random replications of late responses.
TL;DR: In this paper, the authors investigated the effects of different weighting systems (equally, factorial and DEA weights) on the rankings and score distributions of the subjective well-being indicators.
Abstract: There is a growing literature on the assessment of quality of life (QOL) and subjective well-being (SWB) through composite indicators (CI), obtained by aggregating subjective measures of people well-being. Besides the measurement of elementary indicators, the principal challenges in constructing SWB indicator are the aggregation and weighting system. To this respect, literature hasn’t actually reached a unique consensus. The paper investigates the effects that different weighting systems (equally, factorial and DEA weights) have on the rankings and score distributions of the SWB indicators. Data are provide by a sample survey on the quality of life conducted on the residents in the Romagna area during 2010. Results evidence that diverse weighting techniques produce different SWB score distributions while, to a some extent, rankings are maintained.
TL;DR: In this article, a generalized class of estimators using information on two auxiliary va-riables and sampling fraction in simple random sampling was suggested. But the theoretical findings have not been satisfied with an empirical study and the bias and mean squared error for-mulae of suggested class have been derived under large sample approximation and compared with usual unbiased estimator and Singh's (1967) ratio-cumproduct estimator.
Abstract: This paper suggested a generalized class of estimators using information on two auxiliary va-riables and sampling fraction in simple random sampling. The bias and mean squared error for-mulae of suggested class have been derived under large sample approximation and compared with usual unbiased estimator and Singh’s (1967) ratio-cum-product estimator. The theoretical findings have been satisfied with an empirical study.