Statistical Theory Commons^™

A Graph Theoretic Approach To Testing Associations Between Disparate Sources Of Functional Genomic Data, Raji Balasubramanian, Thomas Laframboise, Denise Scholtens, Robert Gentleman

Bioconductor Project Working Papers

The last few years have seen the advent of high-throughput technologies to analyze various properties of the transcriptome and proteome of several organisms. The congruency of these different data sources, or lack thereof, can shed light on the mechanisms that govern cellular function. A central challenge for bioinformatics research is to develop a unified framework for combining the multiple sources of functional genomics information and testing associations between them, thus obtaining a robust and integrated view of the underlying biology.

We present a graph theoretic approach to test the significance of the association between multiple disparate sources of functional genomics …

Semiparametric Regression Analysis Of Mean Residual Life With Censored Survival Data, Ying Qing Chen, Su-Chun Cheng

U.C. Berkeley Division of Biostatistics Working Paper Series

As a function of time t, mean residual life is the remaining life expectancy of a subject given survival up to t. The proportional mean residual life model, proposed by Oakes & Dasu (1990), provides an alternative to the Cox proportional hazards model to study the association between survival times and covariates. In the presence of censoring, we develop semiparametric inference procedures for the regression coefficients of the Oakes-Dasu model using martingale theory for counting processes. We also present simulation studies and an application to the Veterans' Administration lung cancer data.

Nonparametric And Semiparametric Inference For Models Of Tumor Size And Metastasis, Debashis Ghosh

The University of Michigan Department of Biostatistics Working Paper Series

There has been some recent work in the statistical literature for modelling the relationship between the size of primary cancers and the occurrences of metastases. While nonparametric methods have been proposed for estimation of the tumor size distribution at which metastatic transition occurs, their asymptotic properties have not been studied. In addition, no testing or regression methods are available so that potential confounders and prognostic factors can be adjusted for. We develop a unified approach to nonparametric and semiparametric analysis of modelling tumor size-metastasis data in this article. An equivalence between the models considered by previous authors with survival data …

Semiparametic Models And Estimation Procedures For Binormal Roc Curves With Multiple Biomarkers, Debashis Ghosh

The University of Michigan Department of Biostatistics Working Paper Series

In diagnostic medicine, there is great interest in developing strategies for combining biomarkers in order to optimize classification accuracy. A popular model that has been used for receiver operating characteristic (ROC) curve modelling when one biomarker is available is the binormal model. Extension of the model to accommodate multiple biomarkers has not been considered in this literature. Here, we consider a multivariate binormal framework for combining biomarkers using copula functions that leads to a natural multivariate extension of the binormal model. Estimation in this model will be done using rank-based procedures. We show that the Van der Waerden rank score …

On Corrected Score Approach For Proportional Hazards Model With Covariate Measurement Error, Xiao Song, Yijian Huang

UW Biostatistics Working Paper Series

In the presence of covariate measurement error with the proportional hazards model, several functional modeling methods have been proposed. These include the conditional score estimator (Tsiatis and Davidian, 2001), the parametric correction estimator (Nakamura, 1992) and the nonparametric correction estimator (Huang and Wang, 2000, 2003) in the order of weaker assumptions on the error. Although they are all consistent, each suffers from potential difficulties with small samples and substantial measurement error. In this article, upon noting that the conditional score and parametric correction estimators are asymptotically equivalent in the case of normal error, we investigate their relative finite sample performance …

Fundamental Conditions For The Evolution Of Altruism: Towards A Unification Of Theories, Jeffrey Alan Fletcher

Dissertations and Theses

In evolutionary theory the existence of self-sacrificing cooperative traits poses a problem that has engendered decades of debate. The principal theories of the evolution of altruism are inclusive fitness, reciprocal altruism, and multilevel selection. To provide a framework for the unification of these apparently disparate theories, this dissertation identifies two fundamental conditions required for the evolution of altruism: 1) non-zero-sum fitness benefits for cooperation and 2) positive assortment among altruistic behaviors. I demonstrate the underlying similarities in these three theories in the following two ways.

First, I show that the game-theoretic model of the prisoner’s dilemma (PD) is inherent to …

Beta-Normal Distribution: Bimodality Properties And Application, Felix Famoye, Carl Lee, Nicholas Eugene

Journal of Modern Applied Statistical Methods

The beta-normal distribution is characterized by four parameters that jointly describe the location, the scale and the shape properties. The beta-normal distribution can be unimodal or bimodal. This paper studies the bimodality properties of the beta-normal distribution. The region of bimodality in the parameter space is obtained. The beta-normal distribution is applied to fit a numerical bimodal data set. The beta-normal fits are compared with the fits of mixture-normal distribution through simulation.

Some Improvements In Kernel Estimation Using Line Transect Sampling, Omar M. Eidous

Journal of Modern Applied Statistical Methods

Kernel estimation provides a nonparametric estimate of the probability density function from which a set of data is drawn. This article proposes a method to choose a reference density in bandwidth calculation for kernel estimator using line transect sampling. The method based on testing the shoulder condition, if the shoulder condition seems to be valid using as reference the half normal density, while if the shoulder condition does not seem to be valid, we will use exponential reference density. Accordingly, the performances of the resultant estimator are studied under a wide range of underlying models using simulation techniques. The results …

A Visually Adaptive Bayesian Model In Wavelet Regression, Dongfeng Wu

Journal of Modern Applied Statistical Methods

The implementation of a Bayesian approach to wavelet regression that corresponds to the human visual system is examined. Most existing research in this area assumes non-informative priors, that is, a prior with mean zero. A new way is offered to implement prior information that mimics a visual inspection of noisy data, to obtain a first impression about the shape of the function that results in a prior with non-zero mean. This visually adaptive Bayesian (VAB) prior has a simple structure, intuitive interpretation, and is easy to implement. Skorohod topology is suggested as a more appropriate measure in signal recovering than …

Estimation Using Bivariate Extreme Ranked Set Sampling With Application To The Bivariate Normal Distribution, Mohammad Fraiwan Al-Saleh, Hani M. Samawi

Journal of Modern Applied Statistical Methods

In this article, the procedure of bivariate extreme ranked set sampling (BVERSS) is introduced and investigated as a procedure of obtaining more accurate samples for estimating the parameters of bivariate populations. This procedure takes its strength from the advantages of bivariate ranked set sampling (BVRSS) over the usual ranked set sampling in dealing with two characteristics simultaneously, and the advantages of extreme ranked set sampling (ERSS) over usual RSS in reducing the ranking errors and hence in being more applicable. The BVERSS procedure will be applied to the case of the parameters of the bivariate normal distributions. Illustration using real …

Kernel-Based Estimation Of P(X Less Than Y)With Paired Data, Omar M. Eidous, Ayman Baklizi

Journal of Modern Applied Statistical Methods

A point estimation of P(X < Y) was considered. A nonparametric estimator for P(X < Y) was developed using the kernel density estimator of the joint distribution of X and Y, may be dependent. The resulting estimator was found to be similar to the estimator based on the sign statistic, however it assigns smooth continuous scores to each pair of the observations rather than the zero or one scores of the sign statistic. The asymptotic equivalence of the sign statistic and the proposed estimator is shown and a simulation study is conducted to investigate the performance of the proposed estimator. Results indicate that …

Respondent-Generated Intervals (Rgi) For Recall In Sample Surveys, S. James Press

Journal of Modern Applied Statistical Methods

Respondents are asked for both a basic response to a recall-type question, their usage quantity, and are asked to provide lower and upper bounds for the (Respondent-Generated) interval in which their true values might possibly lie. A Bayesian hierarchical model for estimating the population mean and its variance is presented.. Telephone: (989) 774-

Jmasm11: Comparing Two Small Binomial Proportions, James F. Reed Iii

Journal of Modern Applied Statistical Methods

A large volume of research has focused on comparing the difference between two small binomial proportions. Statisticians recognize that Fisher’s Exact test and Yates chi-square test are excessively conservative. Likewise, many statisticians feel that Pearson’s Chi-square or the likelihood statistic may be inappropriate for small samples. Viable alternatives exist.

A Test-Retest Transition Matrix: A Modification Of Mcnemar’S Test, J. Wanzer Drane, W. Gregory Thatcher

Journal of Modern Applied Statistical Methods

McNemar introduced what is known today as a test for symmetry in a two by two contingency tables. The logic of the test is based on a sample of matched pairs with a dichotomous response. In our example, the sample consists of the scores before and after an education program and the responses before and after the program. Each pair of scores is from only one person. The pretest divides the group of responders according to their answers to a dichotomous question. The posttest divides the two groups into two groups of like labels. The result is a two by …

On The Reporting Of Reliability In Content Analysis, Patric R. Spence

Journal of Modern Applied Statistical Methods

This article explores one type of misreporting of reliability that has been seen in recent conference papers and articles using the method of content analysis. The reporting of reliability is central to the validity of claims made using this method. A brief overview of content analysis is offered, followed by the exploration of one type of misreporting of reliability. Suggestions are offered to address the problem.

Validation Studies: Matters Of Dimensionality, Accuracy, And Parsimony With Predictive Discriminant Analysis And Factor Analysis, David A. Walker

Journal of Modern Applied Statistical Methods

Two studies were used as examples that examined issues of dimensionality, accuracy, and parsimony in educational research via the use of predictive discriminant analysis and factor analysis. Using a two-group problem, study 1 looked at how accurately group membership could be predicted from subjects’ test scores. Study 2 looked at the dimensionality structure of an instrument and if it developed constructs that would measure theorized domains.

Stratified Extreme Ranked Set Sample With Application To Ratio Estimators, Hani M. Samawai, Laith J. Saeid

Journal of Modern Applied Statistical Methods

Stratified extreme ranked set sample (SERSS) is introduced. The performance of the combined and separate ratio estimates using SERSS is investigated. Theoretical and simulation study are presented. Results indicate that using SERSS for estimating the ratios is more efficient than using stratified simple random sample (SSRS) and simple random sample (SRS). In some cases it is more efficient than ranked set sample (RSS) and stratified ranked set sample (SRSS), when the underlying distribution is symmetric. An application to real data on the bilirubin level in jaundice babies is introduced to illustrate the method.

Depth Based Permutation Test For General Differences In Two Multivariate Populations, Yonghong Gao

Journal of Modern Applied Statistical Methods

For two p-dimensional data sets, interest exists in testing if they come from the common population distribution. Proposed is a practical, effective and easy to implement procedure for the testing problem. The proposed procedure is a permutation test based on the concept of the depth of one observation relative to some population distribution. The proposed test is demonstrated to be consistent. A small Monte Carlo simulation was conducted to evaluate the power of the proposed test. The proposed test is applied to some numerical examples.

Accurate Binary Decisions For Assessing Coronary Artery Disease, Mehmet Ali Cengiz

Journal of Modern Applied Statistical Methods

Generalized linear models offer convenient and highly applicable tools for modeling and predicting the behavior of random variables in terms of observable factors and covariates. This paper investigates applications of a special case of generalized linear model to improve the accuracy of predictions and decisions adopting Bayesian methods, in the specific context of assessing coronary artery disease. The basic model is developed for this application using binary response. The results clearly demonstrate the potential advantages offered by this approach.

Statistical Pronouncements Iii, Jmasm Editors

Journal of Modern Applied Statistical Methods

No abstract provided.

Meta-Analysis Of Results And Individual Patient Data In Epidemiologal Studies, Aurelio Tobías, Marc Saez, Manolis Kogevinas

Journal of Modern Applied Statistical Methods

Epidemiological information can be aggregated by combining results through a meta-analysis technique, or by pooling and analyzing primary data. Common approaches to analyzing pooled studies through an example on the effect of occupational exposure to wood dust on sinonasal cancer are described. Results were combined applying a meta-analysis technique. Alternatively, primary data from all studies were pooled and re-analyzed using mixed effect models. The combination of individual information rather than results is desirable to facilitate interpretations of epidemiological findings, leading also to more precise estimations and more powerful statistical tests for study heterogeneity.

Multivariate Location: Robust Estimators And Inference, Rand R. Wilcox, H. J. Keselman

Journal of Modern Applied Statistical Methods

The sample mean can have poor efficiency relative to various alternative estimators under arbitrarily small departures from normality. In the multivariate case, (affine equivariant) estimators have been proposed for dealing with this problem, but a comparison of various estimators by Massé and Plante (2003) indicated that the small-sample efficiency of some recently derived methods is rather poor. This article reports that a skipped mean, where outliers are removed via a projection-type outlier detection method, is found to be more satisfactory. The more obvious method for computing a confidence region based on the skipped estimator (using a slight modification of the …

A Power Comparison Of Robust Test Statistics Based On Adaptive Estimators, H. J. Keselman, Rand R. Wilcox, James Algina, Abdul R. Othman

Journal of Modern Applied Statistical Methods

Seven test statistics known to be robust to the combined effects of nonnormality and variance heterogeneity were compared for their sensitivity to detect treatment effects in a one-way completely randomized design containing four groups. The six Welch-James-type heteroscedastic tests adopted either symmetric or asymmetric trimmed means, were transformed for skewness, and used a bootstrap method to assess statistical significance. The remaining test, due to Wilcox and Keselman (2003), used a modification of the well-known one-step M-estimator of central tendency rather than trimmed means. The Welch-James-type test is recommended because for nonnormal data likely to be encountered in applied research settings …

A Rank-Based Estimation Procedure For Linear Models With Clustered Data, Suzanne R. Dubnicka

Journal of Modern Applied Statistical Methods

A rank method is presented for estimating regression parameters in the linear model when observations are correlated. This correlation is accounted for by including a random effect term in the linear model. A method is proposed that makes few assumptions about the random effect and error distribution. The main goal of this article is to determine the distributions for which this method performs well relative to existing methods.

A Generalized Quasi-Likelihood Model Application To Modeling Poverty Of Asian American Women, Jeffrey R. Wilson

Journal of Modern Applied Statistical Methods

A generalized quasi-likelihood function that does not require the assumption of an underlying distribution when modeling jointly the mean and the variance, is introduced to examine poverty of Asian American women living in the West coast of the United States, using data from U.S. Census Bureau.

Estimation Of Multiple Linear Functional Relationships, Amjad D. Al-Nasser

Journal of Modern Applied Statistical Methods

This article deals with multiple linear functional relationships models. Two robust estimations procedure are proposed to estimate the model, based on Generalized Maximum Entropy and Partial Least Square. They are distribution free and do not rely (so much) on classical assumptions. The experiments showed that the GME approach outperforms the PLS in terms of mean squares of errors (MSE). Empirical examples are studied.

Teaching Random Assignment: Do You Believe It Works?, Shlomo S. Sawilowsky

Journal of Modern Applied Statistical Methods

Textbook authors admonish students to check on the comparability of two randomly assigned groups by conducting statistical tests on pretest means to determine if randomization worked. A Monte Carlo study was conducted on a sample of n = 2 per group, where each participant’s personality profile was represented by 7,500 randomly selected and assigned scores. Independent samples t tests were conducted and the results demonstrated that random assignment was successful in equating the two groups on 7,467 variables. The students’ focus is redirected from the ability of random assignment to create comparable groups to the testing of the claims of …

A Comparison Of Bayesian And Frequentist Statistics As Applied In A Simple Repeated Measures Example, Jan Perkins, Daniel Wang

Journal of Modern Applied Statistical Methods

Clinicians see Bayesian and frequentist analysis in published research papers, and need a basic understanding of both. A repeated measures data set was analyzed using both approaches. Assumptions underlying each method and conclusions reached were contrasted. The Bayesian approach is a viable alternative to frequentist statistical analysis for many clinical projects.

Jmasm10: A Fortran Routine For Sieve Bootstrap Prediction Intervals, Andrés M. Alonso

Journal of Modern Applied Statistical Methods

A Fortran routine for constructing nonparametric prediction intervals for a general class of linear processes is described. The approach uses the sieve bootstrap procedure of Bühlmann (1997) based on residual resampling from an autoregressive approximation to the given process.

A Comparison Of Methods For Longitudinal Analysis With Missing Data, James Algina, H. J. Keselman

Journal of Modern Applied Statistical Methods

In a longitudinal two-group randomized trials design, also referred to as randomized parallel-groups design or split-plot repeated measures design, the important hypothesis of interest is whether there are differential rates of change over time, that is, whether there is a group by time interaction. Several analytic methods have been presented in the literature for testing this important hypothesis when data are incomplete. We studied these methods for the case in which the missing data pattern is non-monotone. In agreement with earlier work on monotone missing data patterns, our results on bias, sampling variability, Type I error and power support the …