abstract: Statistical mediation analysis has been widely used in the social sciences in order to examine the indirect effects of an independent variable on a dependent variable. The statistical properties of the single mediator model with manifest and latent variables have been studied using simulation studies. However, the single mediator model with latent variables in the Bayesian framework with various accurate and inaccurate priors for structural and measurement model parameters has yet to be evaluated in a statistical simulation. This dissertation outlines the steps in the estimation of a single mediator model with latent variables as a Bayesian structural equation model (SEM). A Monte Carlo study is carried out in order to examine the statistical properties of point and interval summaries for the mediated effect in the Bayesian latent variable single mediator model with prior distributions with varying degrees of accuracy and informativeness. Bayesian methods with diffuse priors have equally good statistical properties as Maximum Likelihood (ML) and the distribution of the product. With accurate informative priors Bayesian methods can increase power up to 25% and decrease interval width up to 24%. With inaccurate informative priors the point summaries of the mediated effect are more biased than ML estimates, and the bias is higher if the inaccuracy occurs in priors for structural parameters than in priors for measurement model parameters. Findings from the Monte Carlo study are generalizable to Bayesian analyses with priors of the same distributional forms that have comparable amounts of (in)accuracy and informativeness to priors evaluated in the Monte Carlo study. / Dissertation/Thesis / Doctoral Dissertation Psychology 2017

Context: Although the overall prevalence of tuberculosis (TB) in the United States is declining, correctional facilities continue to encounter a higher prevalence of this disease. Despite mandatory reporting laws for active TB, data for latent TB infection (LTBI) remains sketchy because reporting it is not required. Purpose: Investigation of the period prevalence of LTBI in a rural Ohio regional jail compared with other populations in the region to determine the need and adequacy of the screening program. Methods: Data collected on inmates was compared with data collected on hospital employees within the same geographic region. Findings: Between January 2006 and July 2007, staff at the jail tested 1274 inmates for TB using the Mantoux purified protein derivative (PPD) method. Ten inmates (6 in 2006 and 4 in 2007) tested positive. All 10 cases were followed with a negative chest radiograph, leading to the diagnosis of LTBI. The overall incidence for the jail for LTBI was 0.8%, with 0% active cases. However, 85 inmates (6.7% of the population) were released before a PPD interpretation could be completed. In the comparative population, 651 hospital employees were tested for TB. Of these, 32 employees tested positive (LTBI prevalence of 4.9%). There were no cases of active TB reported. Conclusion: The prevalence of LTBI in a rural jail (0.8%) is lower than the comparative sample population at a local hospital (4.9%). The rapid release of inmates (6.7%) indicates that TB data is incomplete and that potential cases of LTBI could have been unreported because of missed opportunity for interpretation of skin tests.

A Monte Carlo simulation study on class enumeration with latent class regression models. / Latent class regression (LCR) is a statistical method used to identify qualitatively different groups or latent classes within a heterogeneous population and commonly used in the behavioural, health, and social sciences. Despite the vast applications, an agreed fit index to correctly determine the number of latent classes is hotly debated. To add, there are also conflicting views on whether covariates should or should not be included into the class enumeration process. We conduct a simulation study to determine the impact of covariates on the class enumeration accuracy as well as study the performance of several commonly used fit indices under different population models and modelling conditions. Our results indicate that of the eight fit indices considered, the aBIC and BLRT proved to be the best performing fit indices for class enumeration. Furthermore, we found that covariates should not be included into the enumeration procedure. Our results illustrate that an unconditional LCA model can enumerate equivalently as well as a conditional LCA model with its true covariate specification. Even with the presence of large covariate effects in the population, the unconditional model is capable of enumerating with high accuracy. As noted by Nylund and Gibson (2016), a misspecified covariate specification can easily lead to an overestimation of latent classes. Therefore, we recommend to perform class enumeration without covariates and determine a set of candidate latent class models with the aBIC. Once that is determined, the BLRT can be utilized on the set of candidate models and confirm whether results obtained by the BLRT match the results of the aBIC. By separating the enumeration procedure of the BLRT, it still allows one to use the BLRT but reduce the heavy computational burden that is associated with this fit index. Subsequent analysis can then be pursued accordingly after the number of latent classes is determined. / Thesis / Master of Science (MSc)

In this thesis, work hardening and latent hardening behaviours of pure Mg single crystals were mainly studied under uniaxial deformation tests at room temperature, 298K. By uniaxial tensile/compression tests, work hardening behaviours of Mg single crystals with different orientations favoured for single and double basal <a> slip, {10-12}<10-11> twin, 2nd order pyramidal <c+a> slip and basal <a> slip + {10-12}<10-11> twin were studied. In order to investigate latent hardening behaviours among slip and twin systems, the Jackson-Basinski type latent hardening experiments in Mg single crystals at room temperature have been carried out under different types of dislocation interactions, which included: (i) the self-interactions, (ii) the co-planar interactions on the basal plane, (iii) basal <a> slip / {10-12}<10-11> twin dislocation interactions, (iv) {10-12}<10-11> twin / basal <a> slip dislocation interactions and (v) basal <a> slip / 2nd order pyramidal <c+a> slip dislocation interactions. The microstructure and micro-texture of the deformed single crystals was observed by optical microscopy (OM), scanning electron microscopy (SEM), and SEM/EBSD methods. In addition, micro- and nano-indentation measurements were performed on adjacent matrix and {10-12}<10-11> twin regions of deformed Mg single crystals and the hardness values were analyzed by the Oliver-Pharr method. The results from the Ph.D. work provided framework for the discussion of the plastic flow in Mg single crystals and quantitative values for hardening parameters used in the crystal plasticity modelling. / Thesis / Doctor of Science (PhD)

No description available.

Cognitive studies of clinical comprehension reveal that expert clinicians are distinguished by their superior ability to recognize meaningful patterns of data in clinical narratives. For example, in psychiatry, the findings of hallucinations and delusions suggest the subdiagnostic hypothesis of a psychotic episode, which in turn suggests several diagnoses, including schizophrenia. This dissertation describes the design and evaluation of a system that aims to simulate an important aspect of expert comprehension: the ability to recognize clusters of findings that support sub-diagnostic hypotheses. The broad range of content in psychiatric narrative presents a formidable barrier to achieving this goal, as it contains general concepts and descriptions of the subjective experience of psychiatric patients in addition to general medical and psychiatric concepts. Lexically driven language processing of such narrative would require the exhaustive predefinition of every concept likely to be encountered. In contrast, Latent Semantic Analysis (LSA) is a corpus-based statistical model of language that learns human-like estimates of the similarity between concepts from a text corpus. LSA is adapted to create trainable models of sub-diagnostic hypotheses, which are then used to recognize related elements in psychiatric discharge summary text. The system is evaluated against an independently annotated set of psychiatric discharge summaries. System-rater agreement approached rater-rater agreement, providing support for the practical application of vector-based models of meaning in domains with broad conceptual territory. Other applications and implications are discussed, including the presentation of a prototype user interface designed to enhance novice comprehension of psychiatric discourse.

Phenolic resins are important components of the composite industry because of their excellent flame retardance and cost effectiveness. However, the common procedure for curing phenolic novolac resins uses hexamethylenetetramine (HMTA) and releases volatiles during the cure, which produce networks with numerous voids. This results in materials that lack the toughness necessary for structural applications. An alternative to curing with HMTA is to crosslink the pendant phenolic groups in the novolac resin with epoxy reagents. This reaction proceeds by nucleophilic addition without the release of any volatiles, thereby creating a void-free network. Flame retardance can be achieved by using an excess of the phenolic component. Network densities can also be controlled to maximize both toughness and stiffness by tailoring the stoichiometry of the reagents. Structure-property relationships of phenolic/epoxy networks have been investigated. Glass transitions decreased, and toughness increased, as the phenolic content in the network was increased. Both results could be correlated to the decrease in network densities along this series, which was investigated by measuring the rubbery moduli well above T_g. Fracture toughness of phenolic/epoxy networks measured by K_1c reached 1.03 MPa-m^1/2, compared with an epoxy control with K^1c = 0.62 MPa-m^1/2 and phenolic control with K_1c = 0.16 MPa-m^1/2. In addition, an increase in novolac content improves flame retardance rather dramatically. The peak heat release rate (PHRR) dropped from 1230 kW/m²⁺ for the epoxy control to 260 kW/m²⁺ for the phenolic/epoxy networks, which approached that of a phenolic resol (PHRR = 116 kW/m²⁺). Phenolic/epoxy composite flame retardance also showed significant improvement when compared to epoxy composites. Melt processability of phenolic/epoxy composites has been achieved through the use of latent nucleophilic initiators. Kinetics of the phenolic/epoxy cure reactions with latent initiators demonstrated that monomeric phosphine initiators yielded faster cure reactions as compared to polymeric initiators. These latent initiators allow composite melt processing, such as prepregging or pultrusion, without premature curing. In addition, cure cycles can be reduced from 4 hours to less than 30 minutes. Composites prepared using these latent initiators had toughness exceeding that of epoxy composites and fatigue limits significantly higher than those of vinyl ester composites. <i>Vita removed, June 10, 2013, per author's request. GMc</i> / Ph. D.

Psychopathy, Machiavellianism, and narcissism are three personality constructs collectively associated with antagonism, callousness, and engagement in socially aversive behaviors. These overlapping personality constructs are theorized to have features that meaningfully distinguish each of them from one another, such as variations in impulsivity. However, investigating the interrelations between existing measures of psychopathy, Machiavellianism, narcissism, and impulsivity presents several methodological challenges. Five Factor Model (FFM) approaches to these four multidimensional constructs provide a promising avenue for examining the associations between these personality domains and impulsivity. This study examined a comprehensive, multidimensional model of impulsivity (i.e., the UPPS-P model of impulsivity) in relation to newly developed FFM-based measures of psychopathy, Machiavellianism, and narcissism using a combination of variable- and person-centered statistical approaches across two distinct samples. Data were analyzed from an archival sample of 918 undergraduate students and 756 MTurk users to provide information on generalizability and replication of results. Hypotheses were tested using a combination of path analyses, latent profile analyses, and multivariate analyses of variance (MANOVA). Results suggest various impulsivity dimensions are associated with underlying aspects of psychopathy, Machiavellianism, and narcissism that may be challenging to delineate when examining domain-level representations of these antagonistic personality constructs. Additionally, results highlight how examining patterns of impulsivity facets may distinguish these personality features. Overall, findings may contribute to a more theoretically precise understanding of how impulsive processes differentiate socially aversive personality features.

Ce travail doctoral a pour premier objectif de replacer les démences dans leur contexte de santé publique en estimant des projections de nombre de cas de démences en France et en Europe jusqu'en 2050. La sensibilité de ces projections aux changements d'hypothèses sur les valeurs d'incidence ou de mortalité des sujets déments, sur le scenario démographique utilisé, et sur la mise en place d'une intervention de prévention, a également été évaluée. Dans ce contexte de forte augmentation du nombre de cas à venir, la prévention des démences, qu'elle soit primaire ou secondaire, sera amenée à tenir une place primordiale dans la prise en charge sociétale de ce problème. Pour pouvoir aboutir à des résultats, les recherches en prévention primaire et secondaire ont besoin de s'appuyer sur une méthodologie adaptée et de sélectionner des critères de jugements pertinents. Le déclin cognitif semble être un critère de jugement de choix, mais son l'utilisation doit éviter un certain nombre d'écueils et de biais. Nous avons dans un premier temps illustré l'analyse de ce critère dans le cadre d'un questionnement de prévention primaire à l'aide d'un modèle non linéaire à variable latente pour données longitudinales. Nous avons pour cela étudié la relation entre consommation chronique de benzodiazépines et déclin cognitif, et montré l'absence d'association sur un large échantillon. Dans un second temps nous avons utilisé ce type de modèle pour décrire et comparer les propriétés métrologiques d'un large ensemble de tests neuropsychologiques dans une cohorte clinique de sujets atteints de déficit cognitif modéré (MCI), et pour étudier la sensibilité de ces tests aux changements cognitifs lié aux prodromes de la maladie d'Alzheimer. Nos travaux ont ainsi permis de fournir des arguments permettant de sélectionner des tests neuropsychologiques susceptibles d'être utilisés dans le cadre de recherches de prévention secondaire pour identifier et/ou suivre les patients présentant un déficit cognitif modéré (MCI) lié à une maladie d'Alzheimer. / The first aim of this doctoral work is to replace dementia in its public health context by estimating the number of dementia cases expected to occur in France and Europe over the next few decades until 2050. The sensitivity of these projections to hypotheses made on dementia incidence and mortality, demographic scenario used, and implementation of a prevention intervention, was also assessed. In this context of increasing number of future cases, the primary and secondary prevention of dementia will take a prominent place in the social management of this problem. Relevant research in the field of primary and secondary prevention requires an appropriate methodology and the use of relevant outcome. Cognitive decline seems to be an appropriate outcome, but a number of biases must be avoided. First, we illustrated the use of this criterion in the context of primary prevention using a nonlinear model with latent variable for longitudinal data to investigated the association between chronic use of benzodiazepines and cognitive decline. We showed the absence of association in a large population-based cohort. Secondly we used this model to describe and compare the metrological properties of a broad range of neuropsychological tests in a clinical cohort of patients with mild cognitive impairment (MCI). We also investigated the sensitivity of these tests to cognitive changes associated with prodromal Alzheimer's disease. Our work provides arguments for selecting neuropsychological tests which can be used in secondary prevention research, to identify and / or to follow patients with mild cognitive impairment (MCI) due to Alzheimer's disease.

In medical, behavioral, and social-psychological sciences, latent variable models are useful in handling variables that cannot be directly measured by a single observed variable, but instead are assessed through a number of observed variables. Traditional latent variable models are usually based on parametric assumptions on both relations between outcome and explanatory latent variables, and error distributions. In this thesis, semiparametric models with Bayesian P-splines are developed to relax these rigid assumptions. / In the fourth part of the thesis, the methodology developed in the third part is further extended to a varying coefficient model with latent variables. Varying coefficient model is a class of flexible semiparametric models in which the effects of covariates are modeled dynamically by unspecified smooth functions. A transformation varying coefficient model can handle arbitrarily distributed dynamic data. A simulation study shows that our proposed method performs well in the analysis of this complex model. / In the last part of the thesis, we propose a finite mixture of varying coefficient models to analyze dynamic data with heterogeneity. A simulation study demonstrates that our proposed method can explore possible existence of different groups in a dynamic data, where in each group the dynamic influences of covariates on the response variables have different patterns. The proposed method is applied to a longitudinal study concerning the effectiveness of heroin treatment. Distinct patterns of heroin use and treatment effect in different patient groups are identified. / In the second part of the thesis, a latent variable model is proposed to relax the first assumption, in which unknown additive functions of latent variables in the structural equation are modeled by Bayesian P-splines. The estimation of nonparametric functions is based on powerful Markov chain Monte Carlo (MCMC) algorithm with block update scheme. A simulation study shows that the proposed method can handle much wider situation than traditional models. The proposed semiparametric latent variable model is applied to a study on osteoporosis prevention and control. Some interesting functional relations, which may be overlooked by traditional parametric latent variable models, are revealed. / In the third part of the thesis, a transformation model is developed to relax the second assumption, which usually assumes the normality of observed variables and random errors. In our proposed model, the nonnormal response variables are transformed to normal by unknown functions modeled with Bayesian P-splines. This semiparametric transformation model is shown to be applicable to a wide range of statistical analysis. The model is applied to a study on the intervention treatment of polydrug use in which the traditional model assumption is violated because many observed variables exhibit serious departure from normality. / Lu, Zhaohua. / Adviser: Xin-Yuan Song. / Source: Dissertation Abstracts International, Volume: 72-04, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (leaves 119-130). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.