for Journals by Title or ISSN
for Articles by Keywords
Followed Journals
Journal you Follow: 0
Sign Up to follow journals, search in your chosen journals and, optionally, receive Email Alerts when new issues of your Followed Journals are published.
Already have an account? Sign In to see the journals you follow.
Journal Cover   International Journal of Advanced Statistics and Probability
  [2 followers]  Follow
  This is an Open Access Journal Open Access journal
   ISSN (Print) 2307-9045 - ISSN (Online) 2307-9045
   Published by Science Publishing Corporation Homepage  [12 journals]
  • An approach to measuring rotatability in central composite designs

    • Authors: Emmanuel Ohaegbulem, Polycarp Chigbu
      Pages: 126 - 131
      Abstract: An approach to measure design rotatability and a measure, that quantifies the percentage of rotatability (from 0 to 100) in the central composite designs are introduced. This new approach is quite different from the ones provided by previous authors which assessed design rotatability by the viewing of tediously obtained contour diagrams. This new approach has not practical limitations, and the measure is very easy to compute. Some examples were used to express this approach.
      PubDate: 2015-05-31
      Issue No: Vol. 3, No. 2 (2015)
  • A reliability assessment of the age-sex data from 1991 and 2006 Nigeria
           population censuses

    • Authors: Emmanuel Ohaegbulem
      Pages: 132 - 137
      Abstract: This study assessed the quality of the age-sex data from 1991 and 2006 Nigeria population censuses using some conventional techniques of evaluating demographic data quality. Whipple and Myers indices were used to determine the extent of digit preference for the age-sex data presented in single years. There were very obvious preference for ages with end-digits 0 and 5 while other end-digits were avoided in the two censuses; and this was more pronounced with the females than males. From 1991 census to 2006 census, the Whipple index for both sexes declined from 294 to 251, while the Myers index declined from 60.8 to 49.9; indications of deficiencies in the Nigeria censuses data. The Joint scores computed for the five-year abridged distribution of the age-sex data yielded 54.83 and 38.52 for the 1991 and 2006 censuses, respectively. This showed that the data are poor in quality and not completely reliable as a result of misreporting. The 29.7% decline in Joint score from 1991 census to 2006 census was an indication of a modest improvement in quality over than fifteen years interval, even as the 2006 age-sex data are usable with adjustment.
      PubDate: 2015-05-31
      Issue No: Vol. 3, No. 2 (2015)
  • Analysis of quantile regression as alternative to ordinary least squares

    • Authors: Ibrahim Abdullahi, Abubakar Yahaya
      Pages: 138 - 145
      Abstract: In this article, an alternative to ordinary least squares (OLS) regression based on analytical solution in the Statgraphics software is considered, and this alternative is no other than quantile regression (QR) model. We also present goodness of fit statistic as well as approximate distributions of the associated test statistics for the parameters. Furthermore, we suggest a goodness of fit statistic called the least absolute deviation (LAD) coefficient of determination. The procedure is well presented, illustrated and validated by a numerical example based on publicly available dataset on fuel consumption in miles per gallon in highway driving.
      PubDate: 2015-06-15
      Issue No: Vol. 3, No. 2 (2015)
  • The complementary Poisson-Lindley class of distributions

    • Authors: Amal Hassan, Salwa Assar, Kareem Ali
      Pages: 146 - 160
      Abstract: This paper proposed a new general class of continuous lifetime distributions, which is a complementary to the Poisson-Lindley family proposed by Asgharzadeh et al. [3]. The new class is derived by compounding the maximum of a random number of independent and identically continuous distributed random variables, and Poisson-Lindley distribution. Several properties of the proposed class are discussed, including a formal proof of probability density, cumulative distribution, and reliability and hazard rate functions. The unknown parameters are estimated by the maximum likelihood method and the Fisher’s information matrix elements are determined. Some sub-models of this class are investigated and studied in some details. Finally, a real data set is analyzed to illustrate the performance of new distributions.
      PubDate: 2015-06-29
      Issue No: Vol. 3, No. 2 (2015)
  • On the asymptotic distribution of an alternative measure of kurtosis

    • Authors: Kagba Suaray
      Pages: 161 - 168
      Abstract: Pearson defined the fourth standardized moment of a symmetric distribution as its kurtosis. There has been much discussion in the literature concerning both the meaning of kurtosis, as well as the effectiveness of the classical sample kurtosis as a reliable measure of peakedness and tail weight. In this paper, we consider an alternative measure, developed by Crow and Siddiqui, used to describe kurtosis. Its value is calculated for a number of common distributions, and a derivation of its asymptotic distribution is given. Simulations follow, which reveal an interesting connection to the literature on the ratio of normal random variables.
      PubDate: 2015-08-07
      Issue No: Vol. 3, No. 2 (2015)
  • An application of two sided power distribution in Bayesian analysis of
           paired comparison of relative importance of predictors in linear
           regression models

    • Authors: Xiaoyin Wang
      Pages: 169 - 184
      Abstract: The purpose of determining the relative importance of predictors is to expose the extent of the individual contribution of a predictor in the presence of other predictors within a selected model. The goal of this article is to expand the current research practice by developing a statistical paired comparison model with Two Sided Power (TSP) link function in the Bayesian framework to evaluate the relative importance of each predictor in a multiple regression model. Results from simulation studies and empirical example reveal that the proposed Two Sided Power link function provides similar conclusions as the commonly used logit link function, but has more advantages from both practical and computational perspectives.
      PubDate: 2015-08-07
      Issue No: Vol. 3, No. 2 (2015)
  • Bayes approach to study scale parameter of log logistic distribution

    • Authors: Wajiha Nasir, Maria Maria, Muhammad Aslam
      Pages: 185 - 190
      Abstract: Scale parameter of Log logistic distribution has been studied using Bayesian approach. Posterior distribution has derived by using non informative prior. Posterior distribution is not in close form so we have work with quadrature numerical integration. Various loss functions has been utilized to derive the Bayes estimators and their corresponding risks. Simulation study has been performed to compare the performance of different estimators.
      PubDate: 2015-09-26
      Issue No: Vol. 3, No. 2 (2015)
  • Missing data treatment method on cluster analysis

    • Authors: Elsiddig Elsadig Mohamed Koko, Amin Ibrahim Adam Mohamed
      Pages: 191 - 209
      Abstract: The missing data in household health survey was challenged for the researcher because of incomplete analysis. The statistical tool cluster analysis methodology implemented in the collected data of Sudan's household health survey in 2006.Current research specifically focuses on the data analysis as the objective is to deal with the missing values in cluster analysis. Two-Step Cluster Analysis is applied in which each participant is classified into one of the identified pattern and the optimal number of classes is determined using SPSS Statistics/IBM. However, the risk of over-fitting of the data must be considered because cluster analysis is a multivariable statistical technique. Any observation with missing data is excluded in the Cluster Analysis because like multi-variable statistical techniques. Therefore, before performing the cluster analysis, missing values will be imputed using multiple imputations (SPSS Statistics/IBM). The clustering results will be displayed in tables. The descriptive statistics and cluster frequencies will be produced for the final cluster model, while the information criterion table will display results for a range of cluster solutions.
      PubDate: 2015-10-18
      Issue No: Vol. 3, No. 2 (2015)
  • How a variable’s partial correlation with other variable(s) can make
           a good predictor: the suppressor variable case

    • Authors: Akinwande Michael Olusegun, Aminu Muktar, Kaile Nasiru Kabir, Ibrahim Abubakar Adamu, Umar Adamu Abubakar
      Pages: 210 - 214
      Abstract: Suppression effect is one of the most elusive and difficult to understand dynamics in multiple regression analysis. Suppressor variable(s) and their dynamics in multiple regression analyses are important in reporting accurate research outcomes. However, quite a number of researchers are unfamiliar with the possible advantages and importance of these variables. Suppressor variables tend to appear useless as separate predictors, but have the potential to change the predictive ability of other variables and completely influence research outcomes. This research describes the role suppressor variables play in a multiple regression analysis and provides practical examples that further explain how suppressor effects can alter research outcomes. Finally, we employed mathematical set notation to demonstrate the concepts of suppressor effects.
      PubDate: 2015-11-05
      Issue No: Vol. 3, No. 2 (2015)
  • Bayesian estimation of the parameters of the bivariate generalized
           exponential distribution using accelerated life testing under censoring

    • Authors: Abd El-Maseh; M. P
      Pages: 215 - 223
      Abstract: In this paper, the Bayesian estimation for the unknown parameters for the bivariate generalized exponential (BVGE) distribution under Bivariate censoring type-I samples with constant stress accelerated life testing (CSALT) are discussed. The scale parameter of the lifetime distribution at constant stress levels is assumed to be an inverse power law function of the stress level. The parameters are estimated by Bayesian approach using Markov Chain Monte Carlo (MCMC) method based on Gibbs sampling. Then, the numerical studies are introduced to illustrate the approach study using samples which have been generated from the BVGE distribution.
      PubDate: 2015-11-12
      Issue No: Vol. 3, No. 2 (2015)
School of Mathematical and Computer Sciences
Heriot-Watt University
Edinburgh, EH14 4AS, UK
Tel: +00 44 (0)131 4513762
Fax: +00 44 (0)131 4513327
About JournalTOCs
News (blog, publications)
JournalTOCs on Twitter   JournalTOCs on Facebook

JournalTOCs © 2009-2015