Statistics and Probability Commons^™

Monte Carlo Simulation Design For Evaluating Normal-Based Control Chart Properties, John N. Dyer

Journal of Modern Applied Statistical Methods

The advent of more complicated control charting schemes has necessitated the use of Monte Carlo simulation (MCS) methods. Unfortunately, few sources exist to study effective design and validation of MCS methods related to control charting. This paper describes the design, issues, considerations and limitations for conducting normal-based control chart MCS studies, including choice of random number generator, simulation size requirements, and accuracy/error in simulation estimation. This paper also describes two design strategies for MCS for control chart evaluations and provides the programming code. As a result, this paper hopes to establish de facto MCS schemes aimed at guiding researchers and …

Latent Variable Model For Weight Gain Prevention Data With Informative Intermittent Missingness, Li Qin, Lisa Weissfeld, Michele Levine, Marsha Marcus, Feng Dai

Journal of Modern Applied Statistical Methods

Missing data is a common problem in longitudinal studies because of the characteristics of repeated measurements. Herein is proposed a latent variable model for nonignorable intermittent missing data in which the latent variables are used as random effects in modeling and link longitudinal responses and missingness process. In this methodology, the latent variables are assumed to be normally distributed with zero-mean, and the values of variance-covariance are calculated through maximum likelihood estimations. Parameter estimates and standard errors of the proposed method are compared with the mixed model and the complete-case analysis in the simulations and the application to the weight …

E-Bayesian Estimation Of The Parameter Of The Logarithmic Series Distribution, Parviz Nasiri, Hassan Esfandyarifar

Journal of Modern Applied Statistical Methods

E-Bayesian estimation is introduced to estimate the parameter of logarithmic series distribution. In addition, E-Bayesian, Bayesian and maximum likelihood estimation with through applying mean squared error.

Jmasm41: An Alternative Method For Multiple Linear Model Regression Modeling, A Technical Combining Of Robust, Bootstrap And Fuzzy Approach (Sas), Wan Muhamad Amir W Ahmad, Mohamad Arif Awang Nawi, Nor Azlida Aleng, Mohamad Shafiq

Journal of Modern Applied Statistical Methods

Research on modeling is becoming popular nowadays, there are several of analyses used in research for modeling and one of them is known as applied multiple linear regressions (MLR). To obtain a bootstrap, robust and fuzzy multiple linear regressions, an experienced researchers should be aware the correct method of statistical analysis in order to get a better improved result. The main idea of bootstrapping is to approximate the entire sampling distribution of some estimator. To achieve this is by resampling from our original sample. In this paper, we emphasized on combining and modeling using bootstrapping, robust and fuzzy regression methodology. …

Misspecification Of Variants Of Autoregressive Garch Models And Effect On In-Sample Forecasting, Olusanya E. Olubusoye, Olaoluwa S. Yaya, Oluwadare O. Ojo

Journal of Modern Applied Statistical Methods

Generally, in empirical financial studies, the determination of the true conditional variance in GARCH modelling is largely subjective. In this paper, we investigate the consequences of choosing a wrong conditional variance specification. The methodology involves specifying a true conditional variance and then simulating data to conform to the true specification. The estimation is then carried out using the true specification and other plausible specification that are appealing to the researcher, using model and forecast evaluation criteria for assessing performance. The results show that GARCH model could serve as better alternative to other asymmetric volatility models.

Improved Ridge Estimator In Linear Regression With Multicollinearity, Heteroscedastic Errors And Outliers, Ashok Vithoba Dorugade

Journal of Modern Applied Statistical Methods

This paper introduces a new estimator, of ridge parameter k for ridge regression and then evaluated by Monte Carlo simulation. We examine the performance of the proposed estimators compared with other well-known estimators for the model with heteroscedastics and/or correlated errors, outlier observations, non-normal errors and suffer from the problem of multicollinearity. It is shown that proposed estimators have a smaller MSE than the ordinary least squared estimator (LS), Hoerl and Kennard (1970) estimator (RR), jackknifed modified ridge (JMR) estimator, and Jackknifed Ridge M‑estimator (JRM).

Multicollinearity And A Ridge Parameter Estimation Approach, Ghadban Khalaf, Mohamed Iguernane

Journal of Modern Applied Statistical Methods

One of the main goals of the multiple linear regression model, Y = Xβ + u, is to assess the importance of independent variables in determining their predictive ability. However, in practical applications, inference about the coefficients of regression can be difficult because the independent variables are correlated and multicollinearity causes instability in the coefficients. A new estimator of ridge regression parameter is proposed and evaluated by simulation techniques in terms of mean squares error (MSE). Results of the simulation study indicate that the suggested estimator dominates ordinary least squares (OLS) estimator and other ridge estimators with respect to …

Estimation Of Parameters Of Misclassified Size Biased Borel Distribution, Bhaktida S. Trivedi, M. N. Patel

Journal of Modern Applied Statistical Methods

A misclassified size-biased Borel Distribution (MSBBD), where some of the observations corresponding to x = c + 1 are wrongly reported as x = c with probability α, is defined. Various estimation methods like the method of maximum likelihood (ML), method of moments, and the Bayes estimation for the parameters of the MSBB distribution are used. The performance of the estimators are studied using simulated bias and simulated risk. Simulation studies are carried out for different values of the parameters and sample size.

Developing Bayesian-Based Confidence Bounds For Non-Identically Distributed Observations Using The Lyapunov Condition, Garry M. Jacyna, Scott L. Rosen

Journal of Modern Applied Statistical Methods

The purpose of this paper is to establish a direct method for assessing the confidence in the detection and identification probabilities for segmented observations that are not identically distributed across assigned segments within a region. This paper arrives at easily computable confidence intervals by showing through mathematical analysis that:

I. The probability of successful detection within each test segment can be characterized by a Beta distribution;
II. The distribution of a weighted sum of independent but non-identically distributed sample means is asymptotically Normally distributed by the Lyapunov variant of the Central Limit Theorem, i.e., the approximation improves as the number …

Jmasm40: Monte Carlo Simulations For Structural Equation Modelling (Revolution R), Sarah A. Rose, Barry Markman

Journal of Modern Applied Statistical Methods

Revolution R code is presented to setup Structural Equation Model (SEM) for a Monte Carlo study. The example is a comparison of different fit indices.

Reflections Concerning Recent Ban On Nhst And Confidence Intervals, Grayson L. Baird, Sunny R. Duerr

Journal of Modern Applied Statistical Methods

This letter addresses some of the immediate consequences of Basic and Applied Social Psychology’s (BASP) ban on null hypothesis significance testing (NHST) and confidence intervals. The letter concludes with three suggestions to improve research in general.

A New Test For Correlation On Bivariate Nonnormal Distributions, Ping Wang, Ping Sa

Journal of Modern Applied Statistical Methods

A new method to conduct a right-tailed test for the correlation on bivariate non-normal distribution is proposed. The comparative simulation study shows that the new test controls the type I error rates well for all the distributions considered. An investigation of the power performance is also provided.

Jmasm42: An Alternative Algorithm And Programming Implementation For Least Absolute Deviation Estimator Of The Linear Regression Models (R), Suraju Olaniyi Ogundele, J. I. Mbegbu, C. R. Nwosu

Journal of Modern Applied Statistical Methods

We propose a least absolute deviation estimation method that produced a least absolute deviation estimator of parameter of the linear regression model. The method is as accurate as existing method.

On Characterizations And Infinite Divisibility Of Recently Introduced Distributions, Gholamhossein G. Hamedani

Mathematics, Statistics and Computer Science Faculty Research and Publications

We present here characterizations of the most recently introduced continuous univariate distributions based on: (i) a simple relationship between two truncated moments; (ii) truncated moments of certain functions of the 1^th order statistic; (iii) truncated moments of certain functions of the n^th order statistic; (iv) truncated moment of certain function of the random variable. We like to mention that the characterization (i) which is expressed in terms of the ratio of truncated moments is stable in the sense of weak convergence. We will also point out that some …

New Classes Of Univariate Continuous Exponential Power Series Distributions, M. Ahsanullah, Gholamhossein G. Hamedani, M. Shakil, B.M. Golam Kibria, F. George

Mathematics, Statistics and Computer Science Faculty Research and Publications

Recently, many researchers have developed various classes of continuous probability distributions which can be generated via the generalized Pearson differential equation and other techniques. In this paper, motivated by the importance of the power series in probability theory and its applications, we derive some new classes of univariate exponential power series distributions for a realvalued continuous random variable, which we call exponential power series distributions. Various mathematical properties of the proposed classes of distributions are discussed. Based on these distributional properties, we have established some characterizations of these distributions as well. It is hoped that the findings of the paper …

Stationary Points For Parametric Stochastic Frontier Models, William C. Horrace, Ian A. Wright

Center for Policy Research

The results of Waldman (1982) on the Normal-Half Normal stochastic frontier model are generalized using the theory of the Dirac delta (Dirac, 1930), and distribution-free conditions are established to ensure a stationary point in the likelihood as the variance of the inefficiency distribution goes to zero. Stability of the stationary point and "wrong skew" results are derived or simulated for common parametric assumptions on the model. Identification is discussed.

A Browser-Based Ide For The Muzecs Platform, Omokolade Hunpatin, Casey O'Hare, Ryan Thomas, Dennis Brylow

Mathematics, Statistics and Computer Science Faculty Research and Publications

We report on a scalable, portable, and secure visual development environment for programming embedded Arduino platforms with Chromebooks in a successful secondary school computer science curriculum. Our web-based environment is part of the larger MUzECS project, an inexpensive replacement module for the Exploring Computer Science (ECS) course being widely deployed in United States high schools. Students use MUzECS to gain a deeper understanding of computing, through a set of blocks which provide appropriate abstractions for working with low-level hardware.

MUzECS improves upon the existing curriculum module by reducing the hardware cost by an order of magnitude, while still preserving the …

Hidden Markov Chain Analysis: Impact Of Misclassification On Effect Of Covariates In Disease Progression And Regression, Haritha Polisetti

USF Tampa Graduate Theses and Dissertations

Most of the chronic diseases have a well-known natural staging system through which the disease progression is interpreted. It is well established that the transition rates from one stage of disease to other stage can be modeled by multi state Markov models. But, it is also well known that the screening systems used to diagnose disease states may subject to error some times. In this study, a simulation study is conducted to illustrate the importance of addressing for misclassification in multi-state Markov models by evaluating and comparing the estimates for the disease progression Markov model with misclassification opposed to disease …

Evaluating The Efficiency Of Treatment Comparison In Crossover Design By Allocating Subjects Based On Ranked Auxiliary Variable, Yisong Huang, Hani Samawi, Robert Vogel, Jingjing Yin, Worlanyo E. Gato, Daniel Linder

Biostatistics: Faculty Publications

The validity of statistical inference depends on proper randomization methods. However, even with proper randomization, we can have imbalanced with respect to important characteristics. In this paper, we introduce a method based on ranked auxiliary variables for treatment allocation in crossover designs using Latin squares models. We evaluate the improvement of the efficiency in treatment comparisons using the proposed method. Our simulation study reveals that our proposed method provides a more powerful test compared to simple randomization with the same sample size. The proposed method is illustrated by conducting an experiment to compare two different concentrations of titanium dioxide nanofiber …

Censoring Unbiased Regression Trees And Ensembles, Jon Arni Steingrimsson, Liqun Diao, Robert L. Strawderman

Johns Hopkins University, Dept. of Biostatistics Working Papers

This paper proposes a novel approach to building regression trees and ensemble learning in survival analysis. By first extending the theory of censoring unbiased transformations, we construct observed data estimators of full data loss functions in cases where responses can be right censored. This theory is used to construct two specific classes of methods for building regression trees and regression ensembles that respectively make use of Buckley-James and doubly robust estimating equations for a given full data risk function. For the particular case of squared error loss, we further show how to implement these algorithms using existing software (e.g., CART, …