Applied Statistics Commons^™

A Comparison Of Statistical Methods Relating Pairwise Distance To A Binary Subject-Level Covariate, Rachael Stone

All Graduate Plan B and other Reports, Spring 1920 to Spring 2023

A community ecologist provided a motivating data set involving a certain animal species with two behavior groups, along with a pairwise genetic distance matrix among individuals. Many community ecologists have analyzed similar data sets with a method known as the Hopkins method, testing for an association between the subject-level covariate (behavior group) and the pairwise distance. This community ecologist wanted to know if they used the Hopkins method, would their results be meaningful? Their question inspired this thesis work, where a different data set was used for confidentiality reasons. Multiple methods (Hopkins method, ADONIS, ANOSIM, and Distance Regression) were used …

Network Exploration Of Correlated Multivariate Protein Data For Alzheimer's Disease Association, Matthew J. Lane

Theses

Alzheimer Disease (AD) is difficult to diagnose by using genetic testing or other traditional methods. Unlike diseases with simple genetic risk components, there exists no single marker determining as to whether someone will develop AD. Furthermore, AD is highly heterogeneous and different subgroups of individuals develop the disease due to differing factors. Traditional diagnostic methods using perceivable cognitive deficiencies are often too little too late due to the brain having suffered damage from decades of disease progression. In order to observe AD at early stages prior to the observation of cognitive deficiencies, biomarkers with greater accuracy are required. By using …

Further Advances For The Sequential Multiple Assignment Randomized Trial (Smart), Tianjiao Dai

Dissertations and Theses (Open Access)

ABSTRACT

FURTHER ADVANCES FOR THE SEQUENTIAL MULTIPLE ASSIGNMENT RANDOMIZED TRIAL (SMART)

Tianjiao Dai, M.S.

Advisory Professor: Sanjay Shete, Ph.D.

Sequential multiple assignment randomized trial (SMART) designs have been developed these years for studying adaptive interventions. In my Ph.D. study, I mainly investigate how to further improve SMART designs and optimize the interventions for each individual in the trial. My dissertation has focused on two topics of SMART designs.

1) Developing a novel SMART design that can reduce the cost and side effects associated with the interventions and proposing the corresponding analytic methods. I have developed a time-varying SMART design in …

Studying The Optimal Scheduling For Controlling Prostate Cancer Under Intermittent Androgen Suppression, Sunil K. Dhar, Hans R. Chaudhry, Bruce G. Bukiet, Zhiming Ji, Nan Gao, Thomas W. Findley

Harvard University Biostatistics Working Paper Series

This retrospective study shows that the majority of patients’ correlations between PSA and Testosterone during the on-treatment period is at least 0.90. Model-based duration calculations to control PSA levels during off-treatment are provided. There are two pairs of models. In one pair, the Generalized Linear Model and Mixed Model are both used to analyze the variability of PSA at the individual patient level by using the variable “Patient ID” as a repeated measure. In the second pair, Patient ID is not used as a repeated measure but additional baseline variables are included to analyze the variability of PSA.

Informational Index And Its Applications In High Dimensional Data, Qingcong Yuan

Theses and Dissertations--Statistics

We introduce a new class of measures for testing independence between two random vectors, which uses expected difference of conditional and marginal characteristic functions. By choosing a particular weight function in the class, we propose a new index for measuring independence and study its property. Two empirical versions are developed, their properties, asymptotics, connection with existing measures and applications are discussed. Implementation and Monte Carlo results are also presented.

We propose a two-stage sufficient variable selections method based on the new index to deal with large p small n data. The method does not require model specification and especially focuses …

Family-Based Association Studies Of Autism In Boys Via Facial-Feature Clusters, Luke Andrew Settles

Masters Theses

"Autism spectrum disorder (ASD) refers to a set of developmental disorders with varied attributes. Due to its substantial heterogeneity in terms of behavioral and clinical phenotypes, it is challenging to discern the genetic biomarkers behind ASD, even though the disease is known to be genetic in nature. This serves as a motivation to detect relationships between single nucleotide polymorphisms (SNPs) and a causal autism disease susceptibility locus (DSL) within more homogeneous subgroups. Recently, clinically meaningful subclassifications of ASD have been discovered utilizing facial features of prepubescent boys. Therefore, through the employment of data from 44 prepubertal Caucasian boys with ASD …

Comparing The Structural Components Variance Estimator And U-Statistics Variance Estimator When Assessing The Difference Between Correlated Aucs With Finite Samples, Anna L. Bosse

Theses and Dissertations

Introduction: The structural components variance estimator proposed by DeLong et al. (1988) is a popular approach used when comparing two correlated AUCs. However, this variance estimator is biased and could be problematic with small sample sizes.

Methods: A U-statistics based variance estimator approach is presented and compared with the structural components variance estimator through a large-scale simulation study under different finite-sample size configurations.

Results: The U-statistics variance estimator was unbiased for the true variance of the difference between correlated AUCs regardless of the sample size and had lower RMSE than the structural components variance estimator, providing better type 1 error …

An Examination Of The Neural Unreliability Thesis Of Autism, John Butler, Sophie Molholm, Gizely Andrade, John J. Foxe

Articles

An emerging neuropathological theory of Autism, referred to here as “the neural unreliability thesis,” proposes greater variability in moment-to-moment cortical representation of environmental events, such that the system shows general instability in its impulse response function. Leading evidence for this thesis derives from functional neuroimaging, a methodology ill-suited for detailed assessment of sensory transmission dynamics occurring at the millisecond scale. Electrophysiological assessments of this thesis, however, are sparse and unconvincing. We conducted detailed examination of visual and somatosensory evoked activity using high-density electrical mapping in individuals with autism (N = 20) and precisely matched neurotypical controls (N = 20), recording …

Gathering Steam In Health Care: A Student History, Michael J. Leach

The Transdisciplinary STEAM+ Journal

In this reflection, I demonstrate STEAM in health care by outlining my 15 years as a university student engaged in formal education, extracurricular learning, research, and employment.

Multilevel Models For Longitudinal Data, Aastha Khatiwada

Electronic Theses and Dissertations

Longitudinal data arise when individuals are measured several times during an ob- servation period and thus the data for each individual are not independent. There are several ways of analyzing longitudinal data when different treatments are com- pared. Multilevel models are used to analyze data that are clustered in some way. In this work, multilevel models are used to analyze longitudinal data from a case study. Results from other more commonly used methods are compared to multilevel models. Also, comparison in output between two software, SAS and R, is done. Finally a method consisting of fitting individual models for each …

Population Projection And Habitat Preference Modeling Of The Endangered James Spinymussel (Pleurobema Collina), Marisa Draper

Senior Honors Projects, 2010-2019

The James Spinymussel (Pleurobema collina) is an endangered mussel species at the top of Virginia’s conservation list. The James Spinymussel plays a critical role in the environment by filtering and cleaning stream water while providing shelter and food for macroinvertebrates; however, conservation efforts are complicated by the mussels’ burrowing behavior, camouflage, and complex life cycle. The goals of the research conducted were to estimate detection probabilities that could be used to predict species presence and facilitate field work, and to track individually marked mussels to test for habitat preferences. Using existing literature and mark-recapture field data, these goals were accomplished …

Identification Of Biomarkers For The Overall Survival Of Ovarian Cancer Patients, Kristi Mai

Graduate Theses and Dissertations

Rapid advance in sequencing technology has led to genome-wide analysis of genetic and epigenetic features simultaneously, making it possible to understand the biological mechanisms underlying cancer initiation and progression. However, how to identify important prognostic features poses a great challenge for both statistical modeling and computing. In this thesis, a network-based approach is applied to the Cancer Genome Atlas (TCGA) ovarian cancer data to identify important genes related to the overall survival of ovarian cancer patients. In the first step, a stepwise correlation-based selector is used to reduce the dimensionality of TCGA data, by filtering out a large number of …

Propensity Score Methods : A Simulation And Case Study Involving Breast Cancer Patients., John Craycroft

Electronic Theses and Dissertations

Observational data presents unique challenges for analysis that are not encountered with experimental data resulting from carefully designed randomized controlled trials. Selection bias and unbalanced treatment assignments can obscure estimations of treatment effects, making the process of causal inference from observational data highly problematic. In 1983, Paul Rosenbaum and Donald Rubin formalized an approach for analyzing observational data that adjusts treatment effect estimates for the set of non-treatment variables that are measured at baseline. The propensity score is the conditional probability of assignment to a treatment group given the covariates. Using this score, one may balance the covariates across treatment …

Strengthening Instrumental Variables Through Weighting, Douglas Lehmann, Yun Li, Rajiv Saran, Yi Li

The University of Michigan Department of Biostatistics Working Paper Series

Instrumental variable (IV) methods are widely used to deal with the issue of unmeasured confounding and are becoming popular in health and medical research. IV models are able to obtain consistent estimates in the presence of unmeasured confounding, but rely on assumptions that are hard to verify and often criticized. An instrument is a variable that influences or encourages individuals toward a particular treatment without directly affecting the outcome. Estimates obtained using instruments with a weak influence over the treatment are known to have larger small-sample bias and to be less robust to the critical IV assumption that the instrument …

Evaluating The Impact Of A Hiv Low-Risk Express Care Task-Shifting Program: A Case Study Of The Targeted Learning Roadmap, Linh Tran, Constantin T. Yiannoutsos, Beverly S. Musick, Kara K. Wools-Kaloustian, Abraham Siika, Sylvester Kimaiyo, Mark J. Van Der Laan, Maya L. Petersen

U.C. Berkeley Division of Biostatistics Working Paper Series

In conducting studies on an exposure of interest, a systematic roadmap should be applied for translating causal questions into statistical analyses and interpreting the results. In this paper we describe an application of one such roadmap applied to estimating the joint effect of both time to availability of a nurse-based triage system (low risk express care (LREC)) and individual enrollment in the program among HIV patients in East Africa. Our study population is comprised of 16;513 subjects found eligible for this task-shifting program within 15 clinics in Kenya between 2006 and 2009, with each clinic starting the LREC program between …

Models For Hsv Shedding Must Account For Two Levels Of Overdispersion, Amalia Magaret

UW Biostatistics Working Paper Series

We have frequently implemented crossover studies to evaluate new therapeutic interventions for genital herpes simplex virus infection. The outcome measured to assess the efficacy of interventions on herpes disease severity is the viral shedding rate, defined as the frequency of detection of HSV on the genital skin and mucosa. We performed a simulation study to ascertain whether our standard model, which we have used previously, was appropriately considering all the necessary features of the shedding data to provide correct inference. We simulated shedding data under our standard, validated assumptions and assessed the ability of 5 different models to reproduce the …

Preparedness Of Hospitals In The Republic Of Ireland For An Influenza Pandemic, An Infection Control Perspective, Mary Reidy, Fiona Ryan, Dervla Hogan, Seán Lacey, Claire Buckley

Department of Mathematics Publications

When an influenza pandemic occurs most of the population is susceptible and attack rates can range as high as 40–50 %. The most important failure in pandemic planning is the lack of standards or guidelines regarding what it means to be ‘prepared’. The aim of this study was to assess the preparedness of acute hospitals in the Republic of Ireland for an influenza pandemic from an infection control perspective.

Zero-Inflated Models To Identify Transcription Factor Binding Sites In Chip-Seq Experiments, Sameera Dhananjaya Viswakula

Mathematics & Statistics Theses & Dissertations

It is essential to determine the protein-DNA binding sites to understand many biological processes. A transcription factor is a particular type of protein that binds to DNA and controls gene regulation in living organisms. Chromatin immunoprecipitation followed by highthroughput sequencing (ChIP-seq) is considered the gold standard in locating these binding sites and programs use to identify DNA-transcription factor binding sites are known as peak-callers. ChIP-seq data are known to exhibit considerable background noise and other biases. In this study, we propose a negative binomial model (NB), a zero-inflated Poisson model (ZIP) and a zero-inflated negative binomial model (ZINB) for peak-calling. …

Proof-Of-Concept Of Environmental Dna Tools For Atlantic Sturgeon Management, Jameson Hinkle

Theses and Dissertations

Abstract

The Atlantic Sturgeon (Acipenser oxyrinchus oxyrinchus, Mitchell) is an anadromous species that spawns in tidal freshwater rivers from Canada to Florida. Overfishing, river sedimentation and alteration of the river bottom have decreased Atlantic Sturgeon populations, and NOAA lists the species as endangered. Ecologists sometimes find it difficult to locate individuals of a species that is rare, endangered or invasive. The need for methods less invasive that can create more resolution of cryptic species presence is necessary. Environmental DNA (eDNA) is a non-invasive means of detecting rare, endangered, or invasive species by isolating nuclear or mitochondrial DNA (mtDNA) from the …

Multi-State Models For Interval Censored Data With Competing Risk, Shaoceng Wei

Theses and Dissertations--Statistics

Multi-state models are often used to evaluate the effect of death as a competing event to the development of dementia in a longitudinal study of the cognitive status of elderly subjects. In this dissertation, both multi-state Markov model and semi-Markov model are used to characterize the flow of subjects from intact cognition to dementia with mild cognitive impairment and global impairment as intervening transient, cognitive states and death as a competing risk.

Firstly, a multi-state Markov model with three transient states: intact cognition, mild cognitive impairment (M.C.I.) and global impairment (G.I.) and one absorbing state: dementia is used to model …

Developments In Nonparametric Regression Methods With Application To Raman Spectroscopy Analysis, Jing Guo

Theses and Dissertations--Epidemiology and Biostatistics

Raman spectroscopy has been successfully employed in the classification of breast pathologies involving basis spectra for chemical constituents of breast tissue and resulted in high sensitivity (94%) and specificity (96%) (Haka et al, 2005). Motivated by recent developments in nonparametric regression, in this work, we adapt stacking, boosting, and dynamic ensemble learning into a nonparametric regression framework with application to Raman spectroscopy analysis for breast cancer diagnosis. In Chapter 2, we apply compound estimation (Charnigo and Srinivasan, 2011) in Raman spectra analysis to classify normal, benign, and malignant breast tissue. We explore both the spectra profiles and their derivatives to …

Meta-Analysis Of Gene Expression Studies, Umaporn Siangphoe

Theses and Dissertations

Combining effect sizes from individual studies using random-effects models are commonly applied in high-dimensional gene expression data. However, unknown study heterogeneity can arise from inconsistency of sample qualities and experimental conditions. High heterogeneity of effect sizes can reduce statistical power of the models. We proposed two new methods for random effects estimation and measurements for model variation and strength of the study heterogeneity. We then developed a statistical technique to test for significance of random effects and identify heterogeneous genes. We also proposed another meta-analytic approach that incorporates informative weights in the random effects meta-analysis models. We compared the proposed …

Stratified Meta-Analysis To Examine Data Biases In Lung Cancer Studies Of Refinery Workers, Sherman Selix

Yale Day of Data

Petroleum refineries employ a variety of workers who historically experienced different potentials for asbestos exposure depending on job tasks. Associations between petroleum refinery work and lung cancer related to occupational asbestos exposure have been quantified among various locations, corporations, and time periods. To combine the data from several individual refinery studies and examine an overall effect, a systematic review and stratified meta-analysis was employed. Using set search terms among four databases, 112 potential publications were identified, of which 29 qualified for meta-analysis. Risk estimates and confidence intervals were extracted from these publications to construct four separate datasets. Inverse variance weighting …

Meta-Analysis Of Social-Personality Psychological Research, Blair T. Johnson, Alice H. Eagly

CHIP Documents

This publication provides a contemporary treatment of the subject of meta-analysis in relation to social-personality psychology. Meta-analysis literally refers to the statistical pooling of the results of independent studies on a given subject, although in practice it refers as well to other steps of research synthesis, including defining the question under investigation, gathering all available research reports, coding of information about the studies and their effects, and interpretation/dissemination of results. Discussed as well are the hallmarks of high-quality meta-analyses.

Genetic Association Testing Of Copy Number Variation, Yinglei Li

Theses and Dissertations--Statistics

Copy-number variation (CNV) has been implicated in many complex diseases. It is of great interest to detect and locate such regions through genetic association testings. However, the association testings are complicated by the fact that CNVs usually span multiple markers and thus such markers are correlated to each other. To overcome the difficulty, it is desirable to pool information across the markers. In this thesis, we propose a kernel-based method for aggregation of marker-level tests, in which first we obtain a bunch of p-values through association tests for every marker and then the association test involving CNV is based on …

Subsemble: An Ensemble Method For Combining Subset-Specific Algorithm Fits, Stephanie Sapp, Mark J. Van Der Laan, John Canny

U.C. Berkeley Division of Biostatistics Working Paper Series

Ensemble methods using the same underlying algorithm trained on different subsets of observations have recently received increased attention as practical prediction tools for massive datasets. We propose Subsemble: a general subset ensemble prediction method, which can be used for small, moderate, or large datasets. Subsemble partitions the full dataset into subsets of observations, fits a specified underlying algorithm on each subset, and uses a clever form of V-fold cross-validation to output a prediction function that combines the subset-specific fits. We give an oracle result that provides a theoretical performance guarantee for Subsemble. Through simulations, we demonstrate that Subsemble can be …

A Bayesian Regression Tree Approach To Identify The Effect Of Nanoparticles Properties On Toxicity Profiles, Cecile Low-Kam, Haiyuan Zhang, Zhaoxia Ji, Tian Xia, Jeffrey I. Zinc, Andre Nel, Donatello Telesca

COBRA Preprint Series

We introduce a Bayesian multiple regression tree model to characterize relationships between physico-chemical properties of nanoparticles and their in-vitro toxicity over multiple doses and times of exposure. Unlike conventional models that rely on data summaries, our model solves the low sample size issue and avoids arbitrary loss of information by combining all measurements from a general exposure experiment across doses, times of exposure, and replicates. The proposed technique integrates Bayesian trees for modeling threshold effects and interactions, and penalized B-splines for dose and time-response surfaces smoothing. The resulting posterior distribution is sampled via a Markov Chain Monte Carlo algorithm. This …

Missing At Random And Ignorability For Inferences About Subsets Of Parameters With Missing Data, Roderick J. Little, Sahar Zanganeh

The University of Michigan Department of Biostatistics Working Paper Series

For likelihood-based inferences from data with missing values, Rubin (1976) showed that the missing data mechanism can be ignored when (a) the missing data are missing at random (MAR), in the sense that missingness does not depend on the missing values after conditioning on the observed data, and (b) the parameters of the data model and the missing-data mechanism are distinct; that is, there are no a priori ties, via parameter space restrictions or prior distributions, between the parameters of the data model and the parameters of the model for the mechanism. Rubin described (a) and (b) as the "weakest …

An Analysis Of Risk Reduction Choices In Dcis Breast Cancer Patients, Lauren Soltesz

Statistics

The main focus of this paper was to evaluate possible demographic and clinical characteristics associated with a woman’s choice of breast conserving surgery (BCS), unilateral mastectomy (ULM), or bilateral risk reduction mastectomy (BRRM). The cohort consisted of patients presenting to the City of Hope National Medical Center with ductal carcinoma in situ breast cancer who elected to have cancer directed surgery (N=305). Analyses to examine associations of patient characteristics with type of surgery were conducted using a multinomial logistic regression. Results showed that older women were more likely to choose breast conserving surgery over bilateral risk reduction mastectomy than younger …

Analyzing Multiple Independent Spatial Point Processes, Neal Grantham

Statistics

No abstract provided.