Ce mémoire porte sur la relation entre les inégalités socioéconomiques et la migration des Indiens inscrits du Québec en 2006. Nous évaluons la contribution des inégalités sociales et scolaires, des inégalités économiques, des inégalités régionales et des inégalités juridiques et politiques à la migration des Indiens inscrits et la contribution de la migration à l’atteinte d’un revenu supérieur. Les résultats obtenus auprès des Autochtones sont comparés à ceux des non-Autochtones de façon à pouvoir distinguer ce qui et spécifique aux Autochtones. Des régressions logistiques ont été effectuées afin d’observer la contribution des caractéristiques individuelles et collectives sur la migration non-récente - entre 2001 et 2005, sur l’atteinte d’un revenu supérieur au 75e centile en 2005 et sur la migration récente - entre 2005 et 2006. Les résultats de la présente recherche montrent que le sexe, l’âge et les variables de migration expliquent très peu le fait de bénéficier d’un revenu supérieur. Obtenir minimalement un diplôme d’études secondaire et d’avoir un travail à temps plein augmente considérablement les chances. Chez les Indiens inscrits, la nation d’appartenance et la zone de résidence expliquent beaucoup plus la variance expliquée. Qui plus est, la contribution de l’âge et du sexe explique très peu le fait d’avoir migré - récemment ou non récemment. Par conséquent, la scolarité, l’occupation expliquent beaucoup le fait que les Indiens inscrits et que les non-Autochtones migrent. Les nations d’appartenance ainsi que les zones de résidence expliquent beaucoup le fait que les Indiens inscrits migrent. / This research shows the relationship between socioeconomics inequalities and migration of the Registered Indians in Quebec in 2006. We evaluate the contribution of social and educational inequalities, economic inequalities, regional inequalities and legal and political inequalities as the contribution of migration to achieve a higher income. The results obtains for the Registered Indians are compared to non-Aboriginals to see the specific situation of the Registered Indians. Logistic regressions are performed to evaluate the contribution of the individual and collective characteristics on non-recent and recent migration and on a 75e percentile income. The results show that sex, age and migration play a very little part in obtaining a higher income. Having minimally a high school diploma or having a full-time job greatly increase the chances. For Registered Indians, nation of belonging and residence area explain more of the variance in income. Furthermore, sex and age explain a very little part of migration recently or less recent. Migration, years of schooling and occupation explain a greater part of the migration of the Registered Indians and non-Aboriginals. Nation of belonging and residence area explains also more of the explained variance of migration.

Cette étude propose d’identifier les facteurs affectant la consommation d’aliments traditionnels à travers une perspective écologique, afin de réduire les taux de prévalence élevés de maladies chroniques et ralentir la forte diminution de consommation d’aliments traditionnels chez les Cris du nord québécois. Pour ce faire, une méthode mixte « sequential explanatory », fut utilisée, combinant quatre groupes focus (n=23) et une régression logistique (n=374) à partir de données secondaires issues de trois études transversales. Selon les résultats de la régression logistique: l’âge, chasser, marcher, le niveau d’éducation et la communauté de résidence étaient associées à une consommation d’aliments traditionnelle trois fois/semaine (p<0,05). Subséquemment, des groupes focus vinrent enrichir et contredire ces résultats. Par exemple : les participants étaient en désaccord avec le fait qu’il n’y avait aucune association entre les aliments traditionnels et l’emploi. Ils croyaient que les personnes sans emploi ont plus d’opportunités pour aller chasser mais peu d’argent pour couvrir les dépenses et inversement pour ceux avec emploi. Ce double effet aurait possiblement fait disparaître l’association dans la régression logistique. Suite aux groupes focus, plusieurs facteurs furent identifiés et distribués dans un modèle écologique suggérant que la consommation d’aliments traditionnels est principalement influencée par des facteurs sociaux, communautaires et environnementaux et ne se limite pas aux facteurs individuels. En conclusion, afin de promouvoir l’alimentation traditionnelle, quatre suggestions de priorités d’action sont proposées. L’alimentation traditionnelle doit faire partie des stratégies de santé publique pour réduire les taux de maladies chroniques et améliorer le bien-être des populations autochtones. / To reduce the high prevalence rates of chronic diseases and slow the sharp decline in consumption of traditional foods for the Cree, northern Quebec, this study proposes to identify the factors affecting the consumption of traditional foods through an ecological perspective. A mixed method design was used combining focus groups (n = 23) with secondary data analysis from three cross-sectional studies (n = 374). First, a logistic regression was performed followed by interviews, providing additional information to enrich, clarify and even contradict interpretations based on quantitative analyzes. In light of the results, two scientific papers were written. The results clearly identified several factors that help or interfer with the consumption of traditional foods. These factors are distributed on four levels of influence :individual, social, community and environmental. A better documentation of environmental, social and community factors is desirable for future research. These findings will have an immediate application in public health planning.

Les modèles incrémentaux sont des modèles statistiques qui ont été développés initialement dans le domaine du marketing. Ils sont composés de deux groupes, un groupe contrôle et un groupe traitement, tous deux comparés par rapport à une variable réponse binaire (le choix de réponses est « oui » ou « non »). Ces modèles ont pour but de détecter l’effet du traitement sur les individus à l’étude. Ces individus n’étant pas tous des clients, nous les appellerons : « prospects ». Cet effet peut être négatif, nul ou positif selon les caractéristiques des individus composants les différents groupes. Ce mémoire a pour objectif de comparer des modèles incrémentaux d’un point de vue bayésien et d’un point de vue fréquentiste. Les modèles incrémentaux utilisés en pratique sont ceux de Lo (2002) et de Lai (2004). Ils sont initialement réalisés d’un point de vue fréquentiste. Ainsi, dans ce mémoire, l’approche bayésienne est utilisée et comparée à l’approche fréquentiste. Les simulations sont e ectuées sur des données générées avec des régressions logistiques. Puis, les paramètres de ces régressions sont estimés avec des simulations Monte-Carlo dans l’approche bayésienne et comparés à ceux obtenus dans l’approche fréquentiste. L’estimation des paramètres a une influence directe sur la capacité du modèle à bien prédire l’effet du traitement sur les individus. Nous considérons l’utilisation de trois lois a priori pour l’estimation des paramètres de façon bayésienne. Elles sont choisies de manière à ce que les lois a priori soient non informatives. Les trois lois utilisées sont les suivantes : la loi bêta transformée, la loi Cauchy et la loi normale. Au cours de l’étude, nous remarquerons que les méthodes bayésiennes ont un réel impact positif sur le ciblage des individus composant les échantillons de petite taille. / Uplift modelling is a statistical method initially developed in marketing. It has two groups (a control group and a treatment group) that are compared using a binary response variable (the response can be « yes » or « no »). The goal of this model is to detect the treatment e ect on prospects. This e ect can be either negative, null or positive. It depends on characteristics of each individual in each group. The purpose of this master thesis is to compare the Bayesian point of view with the frequentist one on uplift modelling. The uplift models used in this thesis are Lo model (2002) and Lai model (2004). Both of them are originally modeled using the frequentist point of view. Therefore, the Bayesian approach is modeled and compared to the frequentist one. Simulations are done on generated data from logistic regressions. Then regression parameters are estimated with Monte- Carlo simulations for Bayesian approach. They are then compared to parameter estimations from the frequentist approach. Parameter estimations have direct influences on the ability of the modelling to predict treatment e ect on individual. Three priors are considered for the Bayesian estimation of the parameters. These densities are chosen such that they are non-informative. They are the following : transformed beta, Cauchy and normal. In the course of the study, we will notice the Bayesian method has a real positive impact on targeting individual from the small size sample.

L'obtention des données a été subventionnée par CIORA (Canadian Initiative for Outcomes in Rheumatology Care). CIORA a aussi financé l'analyse des données effectuées par Justine Zehr. L'Initiative Canadienne Pour Des Resultats En Soins Rhumatologiques (ICORA) a financé l'obtention des données et une partie de l'analyse statistique présentée dans ce mémoire. / La plupart des femmes ayant été atteintes d’arthrite juvénile idiopathique (AJI) continuent de souffrir d’arthrite à l’âge adulte. Certains des médicaments utilisés dans le traitement de l’arthrite tels que les corticostéroïdes et les antiinflammatoires non stéroïdiens (AINS) ne sont pas recommandés durant la grossesse. Le but de ce mémoire est d’estimer l’impact de la prise de ces médicaments sur les événements néfastes à l’accouchement chez ces femmes et leur bébé. Des données administratives sur les prescriptions de médicaments et les hospitalisations d’une cohorte de 1756 femmes ayant souffert d’AJI sont utilisées. Elles ont permis de reconstruire l’historique de consommation de médicaments contre l’arthrite chez les femmes durant la grossesse et l’année précédente. Pour ce faire, deux sous-cohortes de femmes ayant souffert d’AJI ont été formées : une pour la période grossesse et une autre pour la grossesse et l’année précédant celle-ci. Les événements d’intérêt étaient : malformations congénitales, complications néonatales, complications maternelles et petit poids pour l’âge gestationnel. Les proportions de cas présentant l’un de ces événements variaient entre 11,52% et 37,08%. Les médicaments ont été modélisés en terme d’utilisation ou de durée totale de consommation durant la période d’étude. Pour chaque événement, des modèles logistiques ont été estimés pour mesurer l’association entre la prise de médicaments et l’événement, en ajustant pour des variables de confusion potentielles : hypertension avant la grossesse, âge à l’accouchement et obtention du diplôme de secondaire. La consommation de corticostéroïdes semble augmenter statistiquement significativement le risque de présenter des malformations congénitales mais n’avoir aucun impact sur les autres événements. Aucun lien statistiquement significatif n’a été observé entre la consommation de AINS et les événements d’intérêt. / Most women diagnosed with juvenile idiopathic arthritis (JIA) continue to suffer from arthritis in adulthood. Some of the drugs used to treat arthritis such as corticosteroids and non-steroidal anti-inflammatory drugs (NSAIDs) are not recommended during pregnancy. The objective of this thesis is to estimate the impact of these drugs on adverse birth outcomes in women previously diagnosed with JIA and their baby. Administrative data on drug prescriptions and hospitalizations in a cohort of 1756 women with a history of JIA were used to determine individual histories of drug use for the treatment of arthritis during pregnancy and during the year leading to the pregnancy. Two sub-cohorts of women who suffered from JIA were created : one corresponding to the pregnancy and the other to the pregnancy and the year leading to the pregnancy. The events of interest were : congenital anomalies, neonatal adverse outcomes, maternal adverse outcomes and small for gestational age babies. Proportions of the events ranged between 11,52% and 37,08%. Drugs were modelled in terms of use or duration of use during each of the study periods. Logistic regression models were fitted to measure the association between drugs and each of the events, adjusting for the following potential confounding variables : hypertension before pregnancy, maternal age and graduating from high school. The consumption of corticosteroids was associated with a statistically significant increased risk of congenital anomalies but had no impact on the other adverse events. No statistically significant associations were observed between consumption of NSAIDs and the adverse events of interest.

This thesis adapts a life course approach in epidemiology to endometriosis. Endometriosis is a highly prevalent chronic disease affecting women in reproductive age. Firstly, the topic of this disease is introduced, the situation and current knowledge in Australia is discussed. Positive changes in a national level are presented. Secondly, early life exposers and maternal behaviour are investigated as possible risk and protective factors. A systematic review of early life factors identified a low birthweight and formula feeding of infants as risk factors for the development of endometriosis. Lastly, the relation of birthweight, weight at childhood and endometriosis was analysed using data of Australian Longitudinal Study on Women's Health. Other risk and protective factors were evaluated and included into the analysis. Logistic regression was used for determination of statistical significance. High weight at 10 years old was found to be a protective factor against endometriosis.

This study was conducted in the Nkangala district, in the province of Mpumalanga in South Africa. This province remains the largest forestry production region in South Africa. The majority of people living in Mpumalanga are farmers and they have contributed immensely to promote food security. The objective of the study was to determine the level of climate change awareness among small scale maize producers in Mpumalanga province. Random sampling techniques was used to select two hundred and fifty one (251) farmers to be interviewed. A pre-tested questionnaire was administered to maize farmers, focusing on matters relating to climate change awareness in maize production. Data was captured and analysed using software package for social science (SPSS version 20 of 2012). Descriptive statistics were applied to analyse and describe the data. Logistic regression analysis followed to demonstrate the significance of the independent variables on climate change awareness. The results of the analysis indicated that the information received and the size of the farm had an impact on climate change awareness in the area of study. It was therefore recommended that the majority of farmers in Mpumalanga needed to be made aware of climate change in order to assist them to build the adaptive capacity, increase resilience and reduce vulnerability. Information on climate change awareness should be disseminated well to ensure that it will attract the attention of the farmers / Agriculture and  Animal Health / M. Sc. (Agriculture)

En région tempérée, les cartes historiques ont permis de démontrer que l’ancienneté des forêts conditionne les caractéristiques des sols et la présence de certaines espèces végétales. En revanche, peu d’études ont été réalisées pour tester ces différences en région méditerranéenne. L'objectif général de la thèse est d’analyser les déterminants du changement du couvert forestier et l’effet de la continuité temporelle et des usages passés sur les sols et la flore des forêts actuelles en région méditerranéenne. Ce travail prend place au sein du territoire du Parc Naturel Régional du Luberon. Le premier volet identifie les déterminants de la distribution des usages et de la reconquête forestière, à partir de données cartographiques de 1860, 1958 et 2010. Il apparaît notamment que les forêts se sont maintenues sur les terres les moins productives et que la reconquête forestière s’est faite sur des sols peu productifs, et à proximité des forêts préexistantes. Le second volet explore l’effet de la continuité temporelle et des usages passés des forêts sur leurs caractéristiques écologiques, à partir de relevés floristiques et pédologiques. Les forêts anciennes et récentes se répartissent selon la productivité des sols et accueillent des espèces qui se distinguent par leurs traits et leurs exigences écologiques, avec notamment des espèces spécialistes forestières, phanérophytes et endozoochores plus fréquentes en forêt ancienne. Si ce travail s’appuie sur la carte d’État-Major, il s’avère que l’écologie historique des forêts méditerranéennes nécessite également des approches in situ afin d’appréhender la complexité des usages anciens du paysage. / In temperate regions, historical maps have demonstrated that forest ancientness determines soil properties and the presence of some plant species. However, those differences were rarely analysed in the Mediterranean region. The main objectives of this PhD thesis are to analyse the drivers of forest cover change and the effect of temporal continuity and past land uses on forest soils and understory vegetation in the Mediterranean region. This work relies on the territory of the Regional Natural Park of Luberon. The first part consists of identifying biophysical and socioeconomic drivers of land use and forest recovery spatial distribution, based on the comparison of 1860, 1958 and 2010 land uses. Forest maintained on the least productive land while forest recovery occurred on soils with low productivity, and close to pre-existing forests. The second part investigates the effect of forest temporal continuity and past land uses on their ecological characteristics, based on floristic and pedological data. Ancient and recent forests are distributed according to soil productivity and host species of differing traits and ecological preferences. Notably, species significantly preferring ancient forests are true forest species, phanerophytes and endozoochores. If this work relies on the État-Major map, it turns out that the historical ecology of Mediterranean forests requires complementary in situ approaches in order to fully understand the complexity of past landscape uses.

O presente estudo tem o objetivo de analisar os padrões espaciais das ocorrências de suicídio no município de São Paulo, no período de 1996 a 2005, e verificar a sua associação com variávies socioeconômico-culturais (estado civil, renda, instrução, religião, migração). A escolha das variáveis analíticas foi baseada nos fatores de risco ao suicídio levantados pela OMS, OPAS e na teoria sobre o suicídio de Durkheim. Os dados socioeconômicos utilizados foram provenientes do IBGE e os dados de mortalidade do PRO-AIM. Para a identificação do padrão espacial das taxas de suicídio foi utilizado o teste de varredura espacial. Para a verificação da associação com as variáveis socioeconômicas e culturais foi utilizada análise de regressão logísica. No período estudado ocorreram 4275 óbitos por suicídio no município de São Paulo, com uma taxa média de 4,1/100 mil hab/ano. O primeiro teste de varredura espacial, considerando-se 50% da população total como tamanho máximo do agrupamento, identificou 2 agrupamentos significativos, um de risco (RR = 1,66) composto por 18 distritos da região central, centro-sul e centrooeste da cidade (Alto de Pinheiros, Barra Funda, Bela Vista, Bom Retiro, Brás, Cambuci, Consolação, Itaim Bibi, Jardim Paulista, Liberdade, Moema, Morumbi, Pinheiros, Perdizes, República, Santa Cecília, Sé, Vila Mariana) e um de proteção (RR = 0,78) formado por 14 distritos da região sul (Campo Grande, Campo Limpo, Capão Redondo, Cidade Ademar, Cidade Dutra, Grajaú, Jardim Ângela, Jardim São Luís, Santo Amaro, Socorro, Pedreira, Raposo Tavares, Vila Andrade, Vila Sônia). O teste considerando-se 5% da população total como tamanho máximo do agrupamento, encontrou 2 agrupamentos significativos. Nesse teste, o agrupamento de risco do primeiro teste foi desmembrado em dois agrupamentos menores, ambos de risco. O agrupamento primário apresentou RR = 1,92 em 9 distritos centrais, o agrupamento secundário RR = 1,58 em 6 distritos da região centro-sul. Para a análise de regressão logística, o agrupamento de risco identificado no primeiro teste de varredura espacial (18 distritos) e os demais 78 distritos (contraste) foram definidos como variáveis dependentes e as variáveis socioeconômicoculturais independentes. O primeiro modelo ajustado na regressão multivariada identificou as seguintes variáveis como risco: solteiros (OR = 2,36); migrantes (OR = 1,49); católicos (OR = 1,36); elevada renda (OR = 1,05). O segundo modelo multivariado identificou as seguintes variáveis como proteção: casados (OR = 0,48); evangélicos (0,60). Os resultados podem ser explicados pelos fatores de risco da literatura, pela teoria de Durkheim adaptada à realidade paulistana e pela diferença entre suicídio e homicídio. / The present study aims to analyse the space patterns of suicide occurence in the city of São Paulo, in the period from 1996 to 2005, and check its association with the socioeconomic-cultural variables (marital status, income, education, religion, migration). The choice of the analytical variables was based on the suicide risk factors lifted by the WHO and OPAS, and in the Durkheim´s suicide theory. The socioeconomic data used were originated from the IBGE and the mortality data from the PRO-AIM. The spatial scan test was applied to identify space patterns of suicide rates. In order to check the association with the socioeconomic-cultural variables, the logistic regression analysis was used. In the studied period, 4275 suicide deaths took place in the city of São Paulo, with a mean rate of 4,1/100 thousand inhabitant/year. The first spatial scan test, considering 50% of the total population as the maximum cluster size, identified 2 significant clusters, one of risk (RR = 1,66) composed by 18 districts of the central region, south-center and western-center of the city (Alto de Pinheiros, Barra Funda, Bela Vista, Bom Retiro, Brás, Cambuci, Consolação, Itaim Bibi, Jardim Paulista, Liberdade, Moema, Morumbi, Pinheiros, Perdizes, República, Santa Cecília, Sé, Vila Mariana) and one of protection (RR = 0,78) formed by 14 districts of the south region (Campo Grande, Campo Limpo, Capão Redondo, Cidade Ademar, Cidade Dutra, Grajaú, Jardim Ângela, Jardim São Luís, Santo Amaro, Socorro, Pedreira, Raposo Tavares, Vila Andrade, Vila Sônia). The test considering 5% of the total population as the maximum cluster size, found 2 significant clusters. In this test, the risk cluster of the first test was dismembered in two minor clusters, both of risk. The primary cluster presented RR = 1,92 in 9 central districts the secondary cluster RR = 1,58 in 6 districts of the south-centre. To the logistic regression analysis, the risk cluster identified in the first spatial scan test (18 districts) and the others 78 districts (contrast) were defined as dependent variables and the socioeconomic-cultural variables as independent. The first adjusted model in the multivaried regression identified these variables like risk: singles (OR = 2,36); migrant (OR = 1,49); catholics (OR = 1,36); high income (OR = 1,05). The second multivaried model identified these variables like protection: married (OR = 0,48); evangelic (0,60). The results can be explained by the literature risk factors, the Durkheim´s theory well-adjusted for the reality of São Paulo and by the difference between suicide and homicide.

O objetivo do autor neste trabalho foi definir critérios pré-operatórios de invasão do seio cavernoso em imagens de Ressonância Magnética (RM) de pacientes com adenomas hipofisários. Neste estudo retrospectivo, foram revisadas as imagens de RM de 103 pacientes com adenomas hipofisários tratados cirurgicamente (48 com invasão do seio cavernoso) e compararam 8 sinais de imagem com os achados cirúrgicos (critério de referência para invasão). A análise estatística foi realizada utilizando o teste de qui-quadrado (X2), e a sensibilidade, especificidade, valor preditivo positivo(VPP) e valor preditivo negativo (VPN) foram obtidos para cada grupo de sinais. Também foram calculados, por regressão logística, os valores de \"odds ratio\" dos critérios mais significativos, e realizada a regressão logística múltipla para análise conjunta desses critérios. O seio cavernoso não estava invadido com certeza quando: a glândula hipofisária normal estava interposta entre ele e o adenoma (VPP, 100%); o compartimento venoso medial foi visibilizado (VPP, 100%); a porcentagem de envolvimento carotídeo pelo tumor foi menor que 25% (VPN, 100%) e o tumor não cruzava a linha intercarotídea medial (VPN, 100%). A invasão do seio cavernoso era certa (VPP, 100%) se: a porcentagem de envolvimento carotídeo era igual ou maior que 45%; 3 ou mais compartimentos venosos não eram visibilizados e o compartimento venoso lateral não era visibilizado. A presença de invasão era altamente sugestiva quando: o compartimento venoso inferior não era visibilizado (VPP, 92,8%); o tumor cruzava a linha intercarotídea lateral (VPP, 96,1%) e quando a parede dural lateral do seio cavernoso estava abaulada (VPP, 92,3%). Na análise conjunta, o critério com maior significância estatística para invasão foi o envolvimento carotídeo pelo adenoma > 30% / The author\'s objective in this study was to define preoperative MRI criteria of cavernous sinus invasion by pituitary adenoma. In this retrospective study, the authors reviewed the MR images of 103 patients with pituitary adenomas treated surgically (48 with cavernous sinus invasion) and compared 8 groups of MR imaging signs with the surgical findings (the standard of reference criterion for invasion). Statistical analysis was performed using a qui-square test (X2), and the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were obtained for each group of signs. The odds ratio of the most significant criteria was also obtained and the multiple logistic regression test was used to evaluate the criteria all together. The cavernous sinus was definitely not invaded when: normal pituitary gland was interposed between the adenoma and it (PPV, 100%); the medial venous compartment was depicted (PPV, 100%); the percentage of encasement of the intracavernous internal carotid artery (ICA) was lower than 25% (NPV, 100%) and the medial intercarotid line was not crossed (NPV, 100%). Invasion of the cavernous sinus was certain (PPV, 100%) if: the percentage of encasement of the intracavernous ICA was 45% or greater and three or more venous compartment or the lateral venous compartment was not depicted. It was highly probable invaded if: the inferior venous compartment was not depicted (PPV, 92,8%); the lateral intercarotid line was crossed (PPV, 96,1%) and the lateral dural wall of the cavernous sinus was bulged (PPV, 92,3%). The most valuable criterion of cavernous sinus invasion by statistical analysis was the percentage of encasement of intracavernous ICA >30%.

Neste trabalho estudamos o modelo de regressão logística com erro de medida nas covariáveis. Abordamos as metodologias de estimação de máxima pseudoverossimilhança pelo algoritmo EM-Monte Carlo, calibração da regressão, SIMEX e naïve (ingênuo), método este que ignora o erro de medida. Comparamos os métodos em relação à estimação, através do viés e da raiz do erro quadrático médio, e em relação à predição de novas observações, através das medidas de desempenho sensibilidade, especificidade, verdadeiro preditivo positivo, verdadeiro preditivo negativo, acurácia e estatística de Kolmogorov-Smirnov. Os estudos de simulação evidenciam o melhor desempenho do método de máxima pseudoverossimilhança na estimação. Para as medidas de desempenho na predição não há diferença entre os métodos de estimação. Por fim, utilizamos nossos resultados em dois conjuntos de dados reais de diferentes áreas: área médica, cujo objetivo está na estimação da razão de chances, e área financeira, cujo intuito é a predição de novas observações. / We study the logistic model when explanatory variables are measured with error. Three estimation methods are presented, namely maximum pseudo-likelihood obtained through a Monte Carlo expectation-maximization type algorithm, regression calibration, SIMEX and naïve, which ignores the measurement error. These methods are compared through simulation. From the estimation point of view, we compare the different methods by evaluating their biases and root mean square errors. The predictive quality of the methods is evaluated based on sensitivity, specificity, positive and negative predictive values, accuracy and the Kolmogorov-Smirnov statistic. The simulation studies show that the best performing method is the maximum pseudo-likelihood method when the objective is to estimate the parameters. There is no difference among the estimation methods for predictive purposes. The results are illustrated in two real data sets from different application areas: medical area, whose goal is the estimation of the odds ratio, and financial area, whose goal is the prediction of new observations.