Statistical Theory Commons^™

Estimation Of Risk For Developing Cardiac Problem In Patients Of Type 2 Diabetes As Obtained By The Technique Of Density Estimation, Ajit Mukherjee, Ajit Mathur, Rakesh Mittal

Journal of Modern Applied Statistical Methods

High levels of cholesterol and triglyceride are known to be strongly associated with development of cardiac problem in patients of type 2 diabetes. In a hospital-based study, patients showing ECG positive were compared with those who were not. The observations on cholesterol and triglyceride were considered for estimation of risk for developing the cardiac problem. The technique of density estimation employing Epanechnikov kernel was used for estimating bivariate probability density functions with respect to observations on cholesterol and triglyceride of the two groups. Using the odds form of Bayes’ rule, the estimates of posterior odds were computed.

Effects Of Physical Activity On Psychological Change In Advanced Age: A Multivariate Meta-Analysis, Meng-Jia Wu, Betsy Jane Becker, Yael Netz

Journal of Modern Applied Statistical Methods

An example of multivariate meta-analysis is demonstrated by synthesizing the treatment effects of exercise of 15 groups on six mood state changes in elders measured by the Profile of Mood States (POMS) scale. Two different methods were used to analyze this multivariate dataset. The SAS codes for two set of the analyses were provided. Results showed that exercise has a modest and positive impact on elders mood change.

Application Of A New Procedure For Power Analysis And Comparison Of The Adjusted Univariate And Multivariate Tests In Repeated Measures Designs, Sean W. Mulvenon, M. Austin Betz, Kening Wang, Bruno D. Zumbo

Journal of Modern Applied Statistical Methods

A relationship between the multivariate and univariate noncentrality parameters in repeated measures designs was developed for the purpose of assessing the relative power of the univariate and multivariate approaches. An application is provided examining the use of repeated measures designs to evaluate student achievement in a K-12 school system

The Effects Of Heteroscedasticity On Tests Of Equivalence, Jamie A. Gruman, Robert A. Cribbie, Chantal A. Arpin-Cribbie

Journal of Modern Applied Statistical Methods

Tests of equivalence, which are designed to assess the similarity of group means, are becoming more popular, yet very little is known about the statistical properties of these tests. Monte Carlo methods are used to compare the test of equivalence proposed by Schuirmann with modified tests of equivalence that incorporate a heteroscedastic error term. It was found that the latter were more accurate than the Schuirmann test in detecting equivalence when sample sizes and variances were unequal.

On The Properties Of Beta-Gamma Distribution, Lingji Kong, Carl Lee, J.H. Sepanski

Journal of Modern Applied Statistical Methods

A class of generalized gamma distribution called the beta-gamma distribution is proposed. Some of its properties are examined. Its shape can be reversed J-shaped, unimodal, or bimodal. Reliability and hazard functions are also derived, and applications are discussed.

On The Product Of Maxwell And Rice Random Variables, M. Shakil, B. M. Golam Kibria

Journal of Modern Applied Statistical Methods

The distributions of the product of independent random variables arise in many applied problems. These have been extensively studied by many researchers. In this paper, the exact distributions of the product |XY| have been derived when X and Y are Maxwell and Rice random variables respectively, and are distributed independently of each other. The associated cdfs, pdfs, and kth moments have been given.

Optimal Lp-Metric For Minimizing Powered Deviations In Regression, Stan Lipovetsky

Journal of Modern Applied Statistical Methods

Minimizations by least squares or by least absolute deviations are well known criteria in regression modeling. In this work the criterion of generalized mean by powered deviations is suggested. If the parameter of the generalized mean equals one or two, the fitting corresponds to the least absolute or the least squared deviations, respectively. Varying the power parameter yields an optimum value for the objective with a minimum possible residual error. Estimation of a most favorable value of the generalized mean parameter shows that it almost does not depend on data. The optimal power always occurs to be close to 1.7, …

A Comparison Of Two Rank Tests For Repeated Measure Designs, Tian Tian, Rand R. Wilcox

Journal of Modern Applied Statistical Methods

This article compares the small-sample properties of the Agresti-Pendergast and the ATS rank-based method, as described in Brunner, Domh, and Langer (2002), for comparing J dependent groups. The results indicate that the Type I error of the Agresti-Pendergast method is more conservative when J = 2 , but under most conditions, the ATS method performs best in terms of both Type I errors and power.

Jmasm27: An Algorithm For Implementing Gibbs Sampling For 2pno Irt Models (Fortran), Yanyan Sheng, Todd C. Headrick

Journal of Modern Applied Statistical Methods

A Fortran 77 subroutine is provided for implementing the Gibbs sampling procedure to a normal ogive IRT model for binary item response data with the choice of uniform and normal prior distributions for item parameters. The subroutine requires the user to have access to the IMSL library. The source code is available at http://www.siu.edu/~epse1/sheng/Fortran/, along with a stand alone executable file.

Mathmatics In Volume I Of Scripta Universitatis, Shlomo S. Sawilowsky

Journal of Modern Applied Statistical Methods

Immanuel Velikovsky’s journal, Scripta Universitatis, edited by Albert Einstein and first published in 1923, played a significant role in the establishment of the library, and hence, Hebrew University in Jerusalem. The inaugural issue contained an article by the French mathematician Jacques Hadamard. Excerpts from Velikovsky’s diary pertaining to the rationale for the creation of the journal, and the interest in Jewish scholars such as Hadamard, are translated here.

Practical Unit-Root Analysis Using Information Criteria: Simulation Evidence, Kosei Fukuda

Journal of Modern Applied Statistical Methods

The information-criterion-based model selection method for detecting a unit root is proposed. The simulation results suggest that the performances of the proposed method are usually comparable to and sometimes better than those of the conventional unit-root tests. The advantages of the proposed method in practical applications are also discussed.

Tests For Treatment Group Equality When Data Are Nonnormal And Heteroscedastic, Robert A. Cribbie, Rand R. Wilcox, Carmen Bewell, H. J. Keselman

Journal of Modern Applied Statistical Methods

Several tests for group mean equality have been suggested for analyzing nonnormal and heteroscedastic data. A Monte Carlo study compared the Welch tests on ranked data and heterogeneous, nonparametric statistics with previously recommended procedures. Type I error rates for the Welch tests on ranks and the heterogeneous, nonparametric statistics were well controlled with a slight power advantage for the Welch tests on ranks.

A Fano-Huffman Based Statistical Coding Method, Aladdin Shamilov, Senay Asma

Journal of Modern Applied Statistical Methods

Statistical coding techniques have been used for lossless statistical data compression, applying methods such as Ordinary, Shannon, Fano, Enhanced Fano, Huffman and Shannon-Fano-Elias coding methods. A new and improved coding method is presented, the Fano-Huffman Based Statistical Coding Method. It holds the advantages of both the Fano and Huffman coding methods. It is more easily applicable than the Huffman coding methods and it is more optimal than Fano coding method. The optimality with respect to the other methods is realized on the basis of English, German, Turkish, French, Russian and Spanish.

Modeling Longitudinal Ordinal Response Variables For Educational Data, Ann A. O'Connell, Heather Levitt Doucette

Journal of Modern Applied Statistical Methods

This article presents applications for the analysis of multilevel ordinal response data through the proportional odds model. Data are drawn from the public-use Early Childhood Longitudinal Study. Results showed that gender, number of family risk characteristics, and age at kindergarten entry were associated with initial reading proficiency (0 to 5 scale). The number of family risks and age were associated with time-slopes. Three issues are highlighted: building multilevel ordinal models, interpretation of multilevel effects; and determination of predicted probabilities based on results of the multilevel proportional odds models.

Bimodality Revisited, Thomas R. Knapp

Journal of Modern Applied Statistical Methods

Degree of bimodality is an important feature of a frequency distribution, because it could suggest heterogeneity, such as polarization or two underlying distributions combined into one. The literature contains several measures of bimodality. This article attempts to summarize most of those measures, with their attendant advantages and disadvantages.

Type I Error Rates Of The Kenward-Roger Adjusted Degree Of Freedom F-Test For A Split-Plot Design With Missing Values, Miguel A. Padilla, James Algina

Journal of Modern Applied Statistical Methods

The Type I error rate of the Kenward-Roger (KR) test, implemented by PROC MIXED in SAS, was assessed through a simulation study for a one between- and one within-subjects factor split-plot design with ignorable missing values and covariance heterogeneity. The KR test controlled the Type I error well under all of the simulation factors, with all estimated Type I error rates between .040 and .075. The best control was for testing the between-subjects main effect (error rates between .041 and .057) and the worst control was for the between-by-within interaction (.040 to .075). The simulated factors had very small effects …

Comparison Of The T Vs. Wilcoxon Signed-Rank Test For Likert Scale Data And Small Samples, Gary E. Meek, Ceyhun Ozgur, Kenneth Dunning

Journal of Modern Applied Statistical Methods

The one sample t-test is compared with the Wilcoxon Signed-Rank test for identical data sets representing various Likert scales. An empirical approach is used with simulated data. Comparisons are based on observed error rates for 27,850 data sets. Recommendations are provided.

Approximate Bayesian Confidence Intervals For The Mean Of An Exponential Distribution Versus Fisher Matrix Bounds Models, Vincent A. R. Camara

Journal of Modern Applied Statistical Methods

The aim of this article is to obtain and compare confidence intervals for the mean of an exponential distribution. Considering respectively the square error and the Higgins-Tsokos loss functions, approximate Bayesian confidence intervals for parameters of exponential population are derived. Using exponential data, the obtained approximate Bayesian confidence intervals will then be compared to the ones obtained with Fisher Matrix bounds method. It is shown that the proposed approximate Bayesian approach relies only on the observations. The Fisher Matrix bounds method, that uses the z-table, does not always yield the best confidence intervals, and the proposed approach often performs better.

Beta-Weibull Distribution: Some Properties And Applications To Censored Data, Carl Lee, Felix Famoye, Olugbenga Olumolade

Journal of Modern Applied Statistical Methods

Some properties of a four-parameter beta-Weibull distribution are discussed. The beta-Weibull distribution is shown to have bathtub, unimodal, increasing, and decreasing hazard functions. The distribution is applied to censored data sets on bus-motor failures and a censored data set on head-and-neck-cancer clinical trial. A simulation is conducted to compare the beta-Weibull distribution with the exponentiated Weibull distribution.

Covariate Adjustment In Randomized Trials With Binary Outcomes: Targeted Maximum Likelihood Estimation, Kelly L. Moore, Mark J. Van Der Laan

U.C. Berkeley Division of Biostatistics Working Paper Series

Covariate adjustment using linear models for continuous outcomes in randomized trials has been shown to increase efficiency and power over the unadjusted method in estimating the marginal effect of treatment. However, for binary outcomes, investigators generally rely on the unadjusted estimate as the literature indicates that covariate-adjusted estimates based on logistic regression models are less efficient. The crucial step that has been missing when adjusting for covariates is that one must integrate/average the adjusted estimate over those covariates in order to obtain the marginal effect. We apply the method of targeted maximum likelihood estimation (MLE), as presented in van der …

Repeatability Of Serial Carotid Intima Media Thickness Scanning On Individual Subjects, Judi Nightingale

Loma Linda University Electronic Theses, Dissertations & Projects

Background. Heart disease affects millions of Americans each year. In order to improve primary prevention, early risk identification is essential. B-mode ultrasound of the common carotid artery (CCA) intima-media thickness (IMT) assesses risk of a coronary event early in the process of plaque development. Because IMT changes are so small over time, in order to determine clinically significant versus clinically insignificant changes in IMT over a 12 month time period, a protocol is needed which can detect the least significant difference (LSD) of at least 0.030 mm.

Purpose. The purpose of this study was to develop and test a well-defined …

Conservative Estimation Of Optimal Multiple Testing Procedures, James E. Signorovitch

Harvard University Biostatistics Working Paper Series

No abstract provided.

Semiparametric Regression Of Multi-Dimensional Genetic Pathway Data: Least Squares Kernel Machines And Linear Mixed Models, Dawei Liu, Xihong Lin, Debashis Ghosh

Harvard University Biostatistics Working Paper Series

No abstract provided.

Properties Of Monotonic Effects, Tyler J. Vanderweele, James M. Robins

COBRA Preprint Series

Various relationships are shown hold between monotonic effects and weak monotonic effects and the monotonicity of certain conditional expectations. This relationship is considered for both binary and non-binary variables. Counterexamples are provide to show that the results do not hold under less restrictive conditions. The ideas of monotonic effects are furthermore used to relate signed edges on a directed acyclic graph to qualitative effect modification.

Multiple Testing With An Empirical Alternative Hypothesis, James E. Signorovitch

Harvard University Biostatistics Working Paper Series

An optimal multiple testing procedure is identified for linear hypotheses under the general linear model, maximizing the expected number of false null hypotheses rejected at any significance level. The optimal procedure depends on the unknown data-generating distribution, but can be consistently estimated. Drawing information together across many hypotheses, the estimated optimal procedure provides an empirical alternative hypothesis by adapting to underlying patterns of departure from the null. Proposed multiple testing procedures based on the empirical alternative are evaluated through simulations and an application to gene expression microarray data. Compared to a standard multiple testing procedure, it is not unusual for …

Large Cluster Asymptotics For Gee: Working Correlation Models, Hyoju Chung, Thomas Lumley

UW Biostatistics Working Paper Series

This paper presents large cluster asymptotic results for generalized estimating equations. The complexity of working correlation model is characterized in terms of the number of working correlation components to be estimated. When the cluster size is relatively large, we may encounter a situation where a high-dimensional working correlation matrix is modeled and estimated from the data. In the present asymptotic setting, the cluster size and the complexity of working correlation model grow with the number of independent clusters. We show the existence, weak consistency and asymptotic normality of marginal regression parameter estimators using the results of empirical process theory and …

Bayesian Hidden Markov Modeling Of Array Cgh Data, Subharup Guha, Yi Li, Donna Neuberg

Harvard University Biostatistics Working Paper Series

Genomic alterations have been linked to the development and progression of cancer. The technique of Comparative Genomic Hybridization (CGH) yields data consisting of fluorescence intensity ratios of test and reference DNA samples. The intensity ratios provide information about the number of copies in DNA. Practical issues such as the contamination of tumor cells in tissue specimens and normalization errors necessitate the use of statistics for learning about the genomic alterations from array-CGH data. As increasing amounts of array CGH data become available, there is a growing need for automated algorithms for characterizing genomic profiles. Specifically, there is a need for …

Targeted Maximum Likelihood Learning, Mark J. Van Der Laan, Daniel Rubin

U.C. Berkeley Division of Biostatistics Working Paper Series

Suppose one observes a sample of independent and identically distributed observations from a particular data generating distribution. Suppose that one has available an estimate of the density of the data generating distribution such as a maximum likelihood estimator according to a given or data adaptively selected model. Suppose that one is concerned with estimation of a particular pathwise differentiable Euclidean parameter. A substitution estimator evaluating the parameter of the density estimator is typically too biased and might not even converge at the parametric rate: that is, the density estimator was targeted to be a good estimator of the density and …

Spatial Cluster Detection For Censored Outcome Data, Andrea J. Cook, Diane Gold, Yi Li

Harvard University Biostatistics Working Paper Series

No abstract provided.

Diagnosing Bias In The Inverse Probability Of Treatment Weighted Estimator Resulting From Violation Of Experimental Treatment Assignment, Yue Wang, Maya L. Petersen, David Bangsberg, Mark J. Van Der Laan

U.C. Berkeley Division of Biostatistics Working Paper Series

Inverse probability of treatment weighting (IPTW) is frequently used to estimate the causal effects of treatments and interventions. The consistency of the IPTW estimator relies not only on the well-recognized assumption of no unmeasured confounders (Sequential Randomization Assumption or SRA), but also on the assumption of experimentation in the assignment of treatment (Experimental Treatment Assignment or ETA). In finite samples, violations in the ETA assumption can occur due simply to chance; certain treatments become rare or non-existent for certain strata of the population. Such practical violations of the ETA assumption occur frequently in real data, and can result in significant …