This dissertation investigated the optimal strategy for the model specification search in the latent growth modeling. Although developing an initial model based on the theory from prior research is favored, sometimes researchers may need to specify the starting model in the absence of theory. In this simulation study, the effectiveness of the start models in searching for the true population model was examined. The four possible start models adopted in this study were: the simplest mean and covariance structure model, the simplest mean and the most complex covariance structure model, the most complex mean and the simplest covariance structure model, and the most complex mean and covariance structure model. Six model selection criteria were used to determine the recovery of the true model: Likelihood ratio test (LRT), DeltaCFI, DeltaRMSEA, DeltaSRMR, DeltaAIC, and DeltaBIC. The results showed that specifying the most complex covariance structure (UN) with the most complex mean structure recovered the true mean trajectory most successfully with the average hit rate above 90% using the DeltaCFI, DeltaBIC, DeltaAIC, and DeltaSRMR. In searching for the true covariance structure, LRT, DeltaCFI, DeltaAIC, and DeltaBIC performed successfully regardless of the searching method with different start models.

Hydrophilic interaction chromatography is well known as a powerful technique separation of polar and ionizable compound nowadays. However the retention mechanism of the technique is still under debate. Understanding retention mechanism would facilitate the method development using the technique and its future improvement. This was inspiring and became the goal of this thesis. This work involves the characterization of the water enriched layer regarding to water and buffer salt accumulation. Twelve HILIC stationary phase with a diverse surface chemistry regarding to function groups and modification type were studied. Effect of water and salt on regarding to the retention mechanism was investigated by correlating the adsorption data to the retention of selected solutes This also involved the characterization of interactions involve in the separation of 21 HILIC columns. Interactions was probe by retention ratio of pair solutes which are characteristic for each specific interaction. The data was evaluate using principle component analysis – a multivariable data analysis method. The model was comprehensive and its outcomes were confirmed by the studies on adsorptions of water and salts.

The present study investigated the cognitive styles of Indian, Metis, Inuit and non-native adults and adolescents of northern Canada and Alaska. The study identified three relational and two analytical cognitive styles. The styles differed significantly from each other in relation to cultural background, language facility, level of post-secondary education, sex and age of the respondents. Cultural background was found to be the most significant discriminator of those under investigation.<p> Procedure of the study involved the collection of verbalized responses to five open-ended questions concerning education from one hundred northern residents. A total of 528 minutes 32 seconds of taperecorded responses was available from twenty treaty and status Indians, twenty Metis, twenty Inuit and forty non-natives. Subjects included parents, university students, high school students, teacher trainees, teachers, education administrators, native politicians and general community members. The data were submitted to content analysis procedures with items coded according to the Data Analysis of Cognitive Style (DACS) Scale which had been adapted for use in the present study from the work of E. S. Schneidman (1966). Scale item frequencies for each respondent were tabulated and submitted for statistical analyses to the SPSS program discriminant analysis. This analysis identified significantly different functions which translated into patterns of thinking or cognitive styles. In addition this analysis identified the relative importance of functions as discriminators among groups and computed predictability scores which showed the percentage of respondents who were correctly classified according to cognitive styles. and demographic variables.<p>Findings of this study must be considered in relation to the following limitations: the size and nature of the stratified random sample; the reliability of the coders; the use of the unvalidated DACS scale; the ability of the analytical procedures to correctly discriminate among the study groups.<p> The study found that the groups which tended to think in relational styles were: Natives (Indian, Metis, Inuit), people with no university education or with less than one year at university; bilinguals (English and a native language); males; people under twenty years and over forty years of age. The terms Conflict-relational, Moral-relational and Inexactrelational were used to more precisely identify differing cognitive behaviors within the overall relational category. The groups which were found to exhibit analytical cognitive style behaviors included: the nonnative group; those respondents with two to four years of university education; and respondents between thirty and forty years of age. Subcategories within analytical styles were Conflict-analytical and Inexactanalytical.<p>When the Indian, Metis and Inuit respondents were combined into a "native" cultural group they strongly identified with the Moral-relational cognitive style (people-oriented, subjective, holistic, concerned with morals and ethics). The non-native group showed a strong negative relationship to this style. However, when each cultural group was analyzed separately, it was found that the Indian and Inuit subjects were somewhat more analytical (objective, linear, field-independent) than the Metis but less so than the non-natives. On the analysis of four groups, the nonnatives were found to relate to both relational and analytical styles of thinking, indicating a wide range of differences within the group.<p>It was concluded that significant differences existed in the cognitive styles preferred by respondents of different cultural, language, education, sex and age groups in this study. Cultural background was found to be the strongest discriminator in relation to cognitive style differences. It was further concluded that according to extrapolation of findings to the theoretical model it may be possible and desirable to modify curricula content and teaching techniques to achieve a closer match between teaching styles and cognitive and learning styles of. students of indigenous cultural backgrounds.

Genetic data analysis has been capturing a lot of attentions for understanding the mechanism of the development and progressing of diseases like cancers, and is crucial in discovering genetic markers and treatment targets in medical research. This dissertation focuses on several important issues in genetic data analysis, graphical network modeling, feature selection, and covariance estimation. First, we develop a gene network modeling method for discrete gene expression data, produced by technologies such as serial analysis of gene expression and RNA sequencing experiment, which generate counts of mRNA transcripts in cell samples. We propose a generalized linear model to fit the discrete gene expression data and assume that the log ratios of the mean expression levels follow a Gaussian distribution. We derive the gene network structures by selecting covariance matrices of the Gaussian distribution with a hyper-inverse Wishart prior. We incorporate prior network models based on Gene Ontology information, which avails existing biological information on the genes of interest. Next, we consider a variable selection problem, where the variables have natural grouping structures, with application to analysis of chromosomal copy number data. The chromosomal copy number data are produced by molecular inversion probes experiments which measure probe-specific copy number changes. We propose a novel Bayesian variable selection method, the hierarchical structured variable se- lection (HSVS) method, which accounts for the natural gene and probe-within-gene architecture to identify important genes and probes associated with clinically relevant outcomes. We propose the HSVS model for grouped variable selection, where simultaneous selection of both groups and within-group variables is of interest. The HSVS model utilizes a discrete mixture prior distribution for group selection and group-specific Bayesian lasso hierarchies for variable selection within groups. We further provide methods for accounting for serial correlations within groups that incorporate Bayesian fused lasso methods for within-group selection. Finally, we propose a Bayesian method of estimating high-dimensional covariance matrices that can be decomposed into a low rank and sparse component. This covariance structure has a wide range of applications including factor analytical model and random effects model. We model the covariance matrices with the decomposition structure by representing the covariance model in the form of a factor analytic model where the number of latent factors is unknown. We introduce binary indicators for estimating the rank of the low rank component combined with a Bayesian graphical lasso method for estimating the sparse component. We further extend our method to a graphical factor analytic model where the graphical model of the residuals is of interest. We achieve sparse estimation of the inverse covariance of the residuals in the graphical factor model by employing a hyper-inverse Wishart prior method for a decomposable graph and a Bayesian graphical lasso method for an unrestricted graph.

This thesis extends the application of the localized realized volatility model created by Ying Chen, Wolfgang Karl Härdle, and Uta Pigorsch to other futures markets, particularly the CAC 40 and the NI 225. The research attempted to replicate results though ultimately, those results were invalidated by procedural difficulties.

The sharing of needles among injection drug users (IDUs) is a common route of Human Immunodeficiency Virus and Hepatitis C Virus transmission. Through the increased utilization of social network analysis, researchers have been able to examine how the interpersonal relationships of IDUs affect injection risk behaviour. This study involves a secondary analysis of data from a cross-sectional study of 156 IDUs from Winnipeg, Manitoba titled “Social Network Analysis of Injection Drug Users”. Multiple logistic regression analysis was used to assess the individual and the social network characteristics associated with needle sharing among the IDUs. Generalized Estimating Equations analysis was used to determine the injecting dyad characteristics which influence needle sharing behaviour between the IDUs and their injection drug using network members. The results revealed five key thematic findings that were significantly associated with needle sharing: (1) types of drug use, (2) socio-demographic status, (3) injecting in semi-public locations, (4) intimacy, and (5) social influence. The findings from this study suggest that comprehensive prevention approaches that target individuals and their network relationships may be necessary for sustainable reductions in needle sharing among IDUs. / February 2006

The present study investigated the cognitive styles of Indian, Metis, Inuit and non-native adults and adolescents of northern Canada and Alaska. The study identified three relational and two analytical cognitive styles. The styles differed significantly from each other in relation to cultural background, language facility, level of post-secondary education, sex and age of the respondents. Cultural background was found to be the most significant discriminator of those under investigation.<p> Procedure of the study involved the collection of verbalized responses to five open-ended questions concerning education from one hundred northern residents. A total of 528 minutes 32 seconds of taperecorded responses was available from twenty treaty and status Indians, twenty Metis, twenty Inuit and forty non-natives. Subjects included parents, university students, high school students, teacher trainees, teachers, education administrators, native politicians and general community members. The data were submitted to content analysis procedures with items coded according to the Data Analysis of Cognitive Style (DACS) Scale which had been adapted for use in the present study from the work of E. S. Schneidman (1966). Scale item frequencies for each respondent were tabulated and submitted for statistical analyses to the SPSS program discriminant analysis. This analysis identified significantly different functions which translated into patterns of thinking or cognitive styles. In addition this analysis identified the relative importance of functions as discriminators among groups and computed predictability scores which showed the percentage of respondents who were correctly classified according to cognitive styles. and demographic variables.<p>Findings of this study must be considered in relation to the following limitations: the size and nature of the stratified random sample; the reliability of the coders; the use of the unvalidated DACS scale; the ability of the analytical procedures to correctly discriminate among the study groups.<p> The study found that the groups which tended to think in relational styles were: Natives (Indian, Metis, Inuit), people with no university education or with less than one year at university; bilinguals (English and a native language); males; people under twenty years and over forty years of age. The terms Conflict-relational, Moral-relational and Inexactrelational were used to more precisely identify differing cognitive behaviors within the overall relational category. The groups which were found to exhibit analytical cognitive style behaviors included: the nonnative group; those respondents with two to four years of university education; and respondents between thirty and forty years of age. Subcategories within analytical styles were Conflict-analytical and Inexactanalytical.<p>When the Indian, Metis and Inuit respondents were combined into a "native" cultural group they strongly identified with the Moral-relational cognitive style (people-oriented, subjective, holistic, concerned with morals and ethics). The non-native group showed a strong negative relationship to this style. However, when each cultural group was analyzed separately, it was found that the Indian and Inuit subjects were somewhat more analytical (objective, linear, field-independent) than the Metis but less so than the non-natives. On the analysis of four groups, the nonnatives were found to relate to both relational and analytical styles of thinking, indicating a wide range of differences within the group.<p>It was concluded that significant differences existed in the cognitive styles preferred by respondents of different cultural, language, education, sex and age groups in this study. Cultural background was found to be the strongest discriminator in relation to cognitive style differences. It was further concluded that according to extrapolation of findings to the theoretical model it may be possible and desirable to modify curricula content and teaching techniques to achieve a closer match between teaching styles and cognitive and learning styles of. students of indigenous cultural backgrounds.

We introduce a new class of Conley-Morse chain maps for the purpose of comparing the qualitative structure of flows across multiple scales. Conley index theory generalizes classical Morse theory as a tool for studying the dynamics of flows. The qualitative structure of a flow, given a Morse decomposition, can be stored algebraically as a set of homology groups (Conley indices) and a boundary map between the indices (a connection matrix). We show that as long as the qualitative structures of two flows agree on some, perhaps coarse, level we can construct a chain map between the corresponding chain complexes that preserves the relations between the (coarsened) Morse sets. We present elementary examples to motivate applications to data analysis.

In practice today, pressure transient analysis (PTA) and production data analysis (PDA) are done separately and differently by different interpreters in different companies using different analysis techniques, different interpreter-dependent inputs, on pressure and production rate data from the same well, with different software packages. This has led to different analyses outputs and characterizations of the same reservoir. To avoid inconsistent results from different interpretations, this study presents a new way to integrate PTA and PDA on a single diagnostic plot to account for and see the early time and mid-time responses (from the transient tests) and late time (boundary affected/PSS) responses achievable with production analysis, on the same plot; thereby unifying short and long-term analyses and improving the reservoir characterization. The rate normalized pressure (RNP) technique was combined with conventional pressure buildup PTA technique. Data processing algorithms were formulated to improve plot presentation and a stepwise analysis procedure is presented to apply the new technique. The new technique is simple to use and the same conventional interpretation techniques as PTA apply. We have applied the technique to a simulated well case and two field cases. Finally, this new technique represents improvements over previous PDA methods and can help give a long term dynamic description of the well’s drainage area.

Various analytical, semi-analytical, and empirical models have been proposed to characterize rate and pressure behavior as a function of time in tight/shale gas systems featuring a horizontal well with multiple hydraulic fractures. Despite a small number of analytical models and published numerical studies there is currently little consensus regarding the large-scale flow behavior over time in such systems. The purpose of this work is to construct a fit-for-purpose numerical simulator which will account for a variety of production features pertinent to these systems, and to use this model to study the effects of various parameters on flow behavior. Specific features examined in this work include hydraulically fractured horizontal wells, multiple porosity and permeability fields, desorption, and micro-scale flow effects. The theoretical basis of the model is described in Chapter I, along with a validation of the model. We employ the numerical simulator to examine various tight gas and shale gas systems and to illustrate and define the various flow regimes which progressively occur over time. We visualize the flow regimes using both specialized plots of rate and pressure functions, as well as high-resolution maps of pressure distributions. The results of this study are described in Chapter II. We use pressure maps to illustrate the initial linear flow into the hydraulic fractures in a tight gas system, transitioning to compound formation linear flow, and then into elliptical flow. We show that flow behavior is dominated by the fracture configuration due to the extremely low permeability of shale. We also explore the possible effect of microscale flow effects on gas effective permeability and subsequent gas species fractionation. We examine the interaction of sorptive diffusion and Knudsen diffusion. We show that microscale porous media can result in a compositional shift in produced gas concentration without the presence of adsorbed gas. The development and implementation of the micro-flow model is documented in Chapter III. This work expands our understanding of flow behavior in tight gas and shale gas systems, where such an understanding may ultimately be used to estimate reservoir properties and reserves in these types of reservoirs.