A regressão logística com resposta contínua é uma alternativa à regressão logística usual quando a variável resposta possui distribuição contínua e o objetivo do estudo é estimar a probabilidade de ocorrência de valores acima ou abaixo de um determinado valor de corte. O modelo assim construído pode ser escrito na forma de um modelo linear generalizado com função de ligação composta. Quando corretamente especificada, a incorporação da informação sobre a distribuição da variável resposta no modelo faz com que os estimadores de máxima verossimilhança sejam mais eficientes. A técnica é apresentada para os casos em que a variável resposta tem distribuição normal ou log-normal. Como aplicação, considerando dados referentes à cidade de São Paulo nos anos de 1998 e 1999, um modelo de regressão logística com resposta contínua foi considerado na previsão do risco da concentração do poluente NO2 ser maior que um valor de corte estabelecido por legislação. Variáveis climáticas e temporais foram consideradas como preditoras. Mostraram-se importantes para prever o risco a temperatura, a umidade relativa do ar, os dias da semana, as estações do ano, precipitação pluviométrica e velocidade do vento. / Binary regression with continuous outcomes constitutes an alternative to logistic regression when the outcome is continuous and the investigator’s interest focuses to estimate the probability of subjects who fall above or below a cut-off value. The model is based on a generalized linear model with composite link that takes advantage of the continuous structure of the outcome, typically gaussian or lognormal. Under correct response model-ling, binary regression with continuous outcomes is more efficient than logistic regression. A binary regression with continuous outcomes was considered to predict the risk that a NO2 pollutant concentration is above the limits set by environmental legislation in São Paulo city during 1998 and 1999. Climatic and temporal variables were considered as pre-dictors. Temperature, humidity, days of the week, station of the year, precipitation and speed of the wind revealed important to predict the risk.

A regressão logística com resposta contínua é uma alternativa à regressão logística usual quando a variável resposta possui distribuição contínua e o objetivo do estudo é estimar a probabilidade de ocorrência de valores acima ou abaixo de um determinado valor de corte. O modelo assim construído pode ser escrito na forma de um modelo linear generalizado com função de ligação composta. Quando corretamente especificada, a incorporação da informação sobre a distribuição da variável resposta no modelo faz com que os estimadores de máxima verossimilhança sejam mais eficientes. A técnica é apresentada para os casos em que a variável resposta tem distribuição normal ou log-normal. Como aplicação, considerando dados referentes à cidade de São Paulo nos anos de 1998 e 1999, um modelo de regressão logística com resposta contínua foi considerado na previsão do risco da concentração do poluente NO2 ser maior que um valor de corte estabelecido por legislação. Variáveis climáticas e temporais foram consideradas como preditoras. Mostraram-se importantes para prever o risco a temperatura, a umidade relativa do ar, os dias da semana, as estações do ano, precipitação pluviométrica e velocidade do vento. / Binary regression with continuous outcomes constitutes an alternative to logistic regression when the outcome is continuous and the investigator’s interest focuses to estimate the probability of subjects who fall above or below a cut-off value. The model is based on a generalized linear model with composite link that takes advantage of the continuous structure of the outcome, typically gaussian or lognormal. Under correct response model-ling, binary regression with continuous outcomes is more efficient than logistic regression. A binary regression with continuous outcomes was considered to predict the risk that a NO2 pollutant concentration is above the limits set by environmental legislation in São Paulo city during 1998 and 1999. Climatic and temporal variables were considered as pre-dictors. Temperature, humidity, days of the week, station of the year, precipitation and speed of the wind revealed important to predict the risk.

This two-article dissertation summarizes the end-of-course survey and formative feedback literatures, as well as proposes actionability as a useful construct in the analysis of feedback from students captured in real-time during their courses. The present inquiry grew out of my work as the founder of DropThought Education, a Division of DropThought. DropThought Education was a student feedback system that helped instructional designers, instructors, and educational systems to use feedback from students to improve learning and student experience. To find out whether the DropThought style of feedback was more effective than other forms of capturing and analyzing student feedback, I needed to (1) examine the formative feedback literature and (2) test DropThought style feedback against traditional feedback forms. The method and theory proposed demonstrates that feedback from students can be specific and actionable when captured in the moment at students' activity level, in their own words. Application of the real-time feedback approach are relevant to practitioners and researchers alike, whether an instructor looking to improve her class activities, or a learning scientist carrying out interventionist, design-based research.

The Plasmodium vivax Duffy Binding Protein (DBP) is the ligand in the major pathway for P. vivax invasion of human reticulocytes, making it an appealing vaccine candidate. Region II of DBP (DBP-RII) is the minimal portion of the ligand that mediates recognition of the Duffy Antigen Receptor for Chemokines (DARC receptor) on the reticulocyte surface and constitutes the primary vaccine target. Analysis of natural variation in the coding sequences of DBP-RII revealed signature evidence for selective pressure driving variation in the residues of the putative receptor-binding site. We hypothesize that anti-DBP immunity in P. vivax infections is strain-specific and hindered by polymorphic residues altering sensitivity to immune antibody inhibition. To comprehend the human IgG response following P. vivax infections we investigated the specificity of IgG in Pursat Province, Western Cambodia. Using ELISAs, we quantified the antibody titer against five variant alleles of DBP-RII. We also sequenced the DBP-RII of the field isolates to determine their relationship to the variant alleles used in the ELISAs. When correlating the IgG titer between the DBP variants a strain-specific immune response was observed in patients with a high antibody titer to DBP-RII_AH as compared to the other variants. This was different from the correlation of high antibody titers between DBP-RII_P and DBP-RII_7.18 (ρ=0.88, p-value<0.0001) and DBP-RII_P and DBP-RII_O (ρ=0.87, p-value<0.0001). There appeared to be little correlation between specific polymorphic residues and IgG titer. Understanding the immune response to the polymorphisms within PvDBP will allow further identification of epitopes to enable the production of a more effective P. vivax vaccine

The National health and nutrition examination survey (NHANES) is a respected nation-wide program in charge of assessing the health and nutritional status of adults and children in United States. Recent cal research found that folic acid play an important role in preventing baby birth defects. In this paper, we use the generalized estimating equation (GEE) method to study the generalized linear model (GLM) with compound symmetric correlation matrix for the NHANES data and investigate significant factors to ence the intake of food folic acid.

This paper provides an alternative methodology for the analysis of a set of Likert responses measured on a common attitudinal scale when the primary focus of interest is on the relative importance of items in the set. The method makes fewer assumptions about the distribution of the responses than the more usual approaches such as comparisons of means, MANOVA or ordinal data methods. The approach transforms the Likert responses into paired comparison responses between the items. The complete multivariate pattern of responses thus produced can be analysed by an appropriately reformulated paired comparison model. The dependency structure between item responses can also be modelled flexibly. The advantage of this approach is that sets of Likert responses can be analysed simultaneously within the Generalized Linear Model framework, providing standard likelihood based inference for model selection. This method is applied to a recent international survey on the importance of environmental problems. (author's abstract) / Series: Research Report Series / Department of Statistics and Mathematics

This thesis develops hierarchical spatial models for the analysis of correlated and overdispersed count data based on the negative binomial distribution. Model development is motivated by a large scale study of fire frequency in British Columbia, conducted by the Pacific Forestry Service. Specific to our analysis, the main focus lies in examining the interaction between wildfire and forest insect outbreaks. In particular, we wish to relate the frequency of wildfire to the severity of mountain pine beetle (MPB) outbreaks in the province. There is a widespread belief that forest insect outbreaks lead to an increased frequency of wildfires; however, empirical evidence to date has been limited and thus a greater understanding of the association is required. This is critically important as British Columbia is currently experiencing a historically unprecedented MPB outbreak. We specify regression models for fire frequency incorporating random effects in a generalized linear mixed modeling framework. Within such a framework, both spatial correlation and extra-Poisson variation can be accommodated through random effects that are incorporated into the linear predictor of a generalized linear model. We consider a range of models, and conduct model selection and inference within the Bayesian framework with implementation based on Markov Chain Monte Carlo.

Dados binários ou dicotômicos são comuns em muitas áreas das ciências, nas quais, muitas vezes, há interesse em registrar a ocorrência, ou não, de um evento particular. Por outro lado, quando cada unidade amostral é avaliada em mais de uma ocasião no tempo, tem-se dados longitudinais ou medidas repetidas no tempo. é comum também, nesses estudos, se ter uma ou mais variáveis explicativas associadas às variáveis respostas. As variáveis explicativas podem ser dependentes ou independentes do tempo. Na literatura, há técnicas disponíveis para a modelagem e análise desses dados, sendo os modelos disponíveis extensões dos modelos lineares generalizados. O enfoque do presente trabalho é dado aos modelos lineares generalizados de transição para a análise de dados longitudinais envolvendo uma resposta do tipo binária. Esses modelos são baseados em processos estocásticos e o interesse está em modelar as probabilidades de mudanças ou transições de categorias de respostas dos indivíduos no tempo. A suposição mais utilizada nesses processos é a da propriedade markoviana, a qual condiciona a resposta numa dada ocasião ao estado na ocasião anterior. Assim, são revistos os fundamentos para se especificar tais modelos, distinguindo-se os casos estacionário e não-estacionário. O método da máxima verossimilhança é utilizado para o ajuste dos modelos e estimação das probabilidades. Adicionalmente, apresentam-se testes assintóticos para comparar tratamentos, baseados na razão de chances e na diferença das probabilidades de transição. Outra questão explorada é a combinação do modelo de efeitos aleatórios com a do modelo de transição. Os métodos são ilustrados com um exemplo da área da saúde. Para esses dados, o processo é considerado estacionário de ordem dois e o teste proposto sinaliza diferença estatisticamente significativa a favor do tratamento ativo. Apesar de ser uma abordagem inicial dessa metodologia, verifica-se, que os modelos de transição têm notável aplicabilidade e são fontes para estudos e pesquisas futuras. / Binary or dichotomous data are quite common in many fields of Sciences in which there is an interest in registering the occurrence of a particular event. On the other hand, when each sampled unit is evaluated in more than one occasion, we have longitudinal data or repeated measures over time. It is also common, in longitudinal studies, to have explanatory variables associated to response measures, which can be time dependent or independent. In the literature, there are many approaches to modeling and evaluating these data, where the models are extensions of generalized linear models. This work focus on generalized linear transition models suitable for analyzing longitudinal data with binary response. Such models are based on stochastic processes and we aim to model the probabilities of change or transitions of individual response categories in time. The most used assumption in these processes is the Markov property, in which the response in one occasion depends on the immediately preceding response. Thus we review the fundamentals to specify these models, showing the diferences between stationary and non-stationary processes. The maximum likelihood approach is used in order to fit the models and estimate the probabilities. Furthermore, we show asymptotic tests to compare treatments based on odds ratio and on the diferences of transition probabilities. We also present a combination of random-efects model with transition model. The methods are illustrated with health data. For these data, the process is stationary of order two and the suggested test points to a significant statistical diference in favor of the active treatment. This work is an initial approach to transition models, which have high applicability and are great sources for further studies and researches.

In non-life insurance, the independence between the number and size of claims is usually assumed. However, this thesis shows that the assumption of independence can be omitted. We deal with the dependency modeling between frequency and severity of claims. For including the dependence to the total claims model, we consider two methods. The first method uses generalized linear models and the second method used in the thesis is based on dependence modeling by copulas. We also perform a model with independent frequency and severity of claims. This model is compared with the described methods in the simulation part of the thesis. We include dependency on explanatory (rating) variables in all of these models. 1

Les données à haut-débit, par leur grande dimension et leur hétérogénéité, ont motivé le développement de méthodes statistiques pour la sélection de variables. En effet, le signal est souvent observé simultanément à plusieurs facteurs de confusion. Les approches de sélection habituelles, construites sous l'hypothèse d'indépendance des variables, sont alors remises en question car elles peuvent conduire à des décisions erronées. L'objectif de cette thèse est de contribuer à l'amélioration des méthodes de sélection de variables pour la régression et la classification supervisée, par une meilleure prise en compte de la dépendance entre les statistiques de sélection. L'ensemble des méthodes proposées s'appuie sur la description de la dépendance entre covariables par un petit nombre de variables latentes. Ce modèle à facteurs suppose que les covariables sont indépendantes conditionnellement à un vecteur de facteurs latents. Une partie de ce travail de thèse porte sur l'analyse de données de potentiels évoqués (ERP). Les ERP sont utilisés pour décrire par électro-encéphalographie l'évolution temporelle de l'activité cérébrale. Sur les courts intervalles de temps durant lesquels les variations d'ERPs peuvent être liées à des conditions expérimentales, le signal psychologique est faible, au regard de la forte variabilité inter-individuelle des courbes ERP. En effet, ces données sont caractérisées par une structure de dépendance temporelle forte et complexe. L'analyse statistique de ces données revient à tester pour chaque instant un lien entre l'activité cérébrale et des conditions expérimentales. Une méthode de décorrélation des statistiques de test est proposée, basée sur la modélisation jointe du signal et de la dépendance à partir d'une connaissance préalable d'instants où le signal est nul. Ensuite, l'apport du modèle à facteurs dans le cadre général de l'Analyse Discriminante Linéaire est étudié. On démontre que la règle linéaire de classification optimale conditionnelle aux facteurs latents est plus performante que la règle non-conditionnelle. Un algorithme de type EM pour l'estimation des paramètres du modèle est proposé. La méthode de décorrélation des données ainsi définie est compatible avec un objectif de prédiction. Enfin, on aborde de manière plus formelle les problématiques de détection et d'identification de signal en situation de dépendance. On s'intéresse plus particulièrement au Higher Criticism (HC), défini sous l'hypothèse d'un signal rare de faible amplitude et sous l'indépendance. Il est montré dans la littérature que cette méthode atteint des bornes théoriques de détection. Les propriétés du HC en situation de dépendance sont étudiées et les bornes de détectabilité et d'estimabilité sont étendues à des situations arbitrairement complexes de dépendance. Dans le cadre de l'identification de signal, une adaptation de la méthode Higher Criticism Thresholding par décorrélation par les innovations est proposée. / The analysis of high throughput data has renewed the statistical methodology for feature selection. Such data are both characterized by their high dimension and their heterogeneity, as the true signal and several confusing factors are often observed at the same time. In such a framework, the usual statistical approaches are questioned and can lead to misleading decisions as they are initially designed under independence assumption among variables. The goal of this thesis is to contribute to the improvement of variable selection methods in regression and supervised classification issues, by accounting for the dependence between selection statistics. All the methods proposed in this thesis are based on a factor model of covariates, which assumes that variables are conditionally independent given a vector of latent variables. A part of this thesis focuses on the analysis of event-related potentials data (ERP). ERPs are now widely collected in psychological research to determine the time courses of mental events. In the significant analysis of the relationships between event-related potentials and experimental covariates, the psychological signal is often both rare, since it only occurs on short intervals and weak, regarding the huge between-subject variability of ERP curves. Indeed, this data is characterized by a temporal dependence pattern both strong and complex. Moreover, studying the effect of experimental condition on brain activity for each instant is a multiple testing issue. We propose to decorrelate the test statistics by a joint modeling of the signal and time-dependence among test statistics from a prior knowledge of time points during which the signal is null. Second, an extension of decorrelation methods is proposed in order to handle a variable selection issue in the linear supervised classification models framework. The contribution of factor model assumption in the general framework of Linear Discriminant Analysis is studied. It is shown that the optimal linear classification rule conditionally to these factors is more efficient than the non-conditional rule. Next, an Expectation-Maximization algorithm for the estimation of the model parameters is proposed. This method of data decorrelation is compatible with a prediction purpose. At last, the issues of detection and identification of a signal when features are dependent are addressed more analytically. We focus on the Higher Criticism (HC) procedure, defined under the assumptions of a sparse signal of low amplitude and independence among tests. It is shown in the literature that this method reaches theoretical bounds of detection. Properties of HC under dependence are studied and the bounds of detectability and estimability are extended to arbitrarily complex situations of dependence. Finally, in the context of signal identification, an extension of Higher Criticism Thresholding based on innovations is proposed.