By improving the precision and accuracy of public health surveillance tools, we can improve cost-efficacy and obtain meaningful information to act upon. In this dissertation, we propose statistical methods for improving public health surveillance research. In Chapter 1, we introduce a pooled testing option for HIV prevalence estimation surveys to increase testing consent rates and subsequently decrease non-response bias. Pooled testing is less certain than individual testing, but, if more people to submit to testing, then it should reduce the potential for non-response bias. In Chapter 2, we illustrate technical issues in the design of neonatal tetanus elimination surveys. We address identifying the target population; using binary classification via lot quality assurance sampling (LQAS); and adjusting the design for the sensitivity of the survey instrument. In Chapter 3, we extend LQAS survey designs for monitoring malnutrition for longitudinal surveillance programs. By combining historical information with data from previous surveys, we detect spikes in malnutrition rates. Using this framework, we detect rises in malnutrition prevalence in longitudinal programs in Kenya and the Sudan. In Chapter 4, we develop a computationally efficient geostatistical disease mapping model that naturally handles model fitting issues due to temporal boundary misalignment by assuming that an underlying continuous risk surface induces spatial correlation between areas. We apply our method to assess socioeconomic trends in breast cancer incidence in Los Angeles between 1990 and 2000. In Chapter 5, we develop a statistical framework for addressing statistical uncertainty associated with denominator interpolation and with temporal misalignment in disease mapping studies. We propose methods for assessing the impact of the uncertainty in these predictions on health effects analyses. Then, we construct a general framework for spatial misalignment in regression.

The literature is clear that intentional or unintentional clustering of data elements typically results in the inflation of the estimated standard error of fixed parameter estimates. This study is unique in that it examines the impact of multilevel data structures on subject ability which are random effect predictions known as empirical Bayes estimates in the one-parameter IRT / Rasch model. The literature on the impact of complex survey design on latent trait models is mixed and there is no "best practice" established regarding how to handle this situation. A simulation study was conducted to address two questions related to ability estimation. First, what impacts does design based clustering have with respect to desirable statistical properties when estimating subject ability with the one-parameter IRT / Rasch model? Second, since empirical Bayes estimators have shrinkage properties, what impacts does clustering of first-stage sampling units have on measurement validity-does the first-stage sampling unit impact the ability estimate, and if so, is this desirable and equitable? Two models were fit to a factorial experimental design where the data were simulated over various conditions. The first model Rasch model formulated as a HGLM ignores the sample design (incorrect model) while the second incorporates a first-stage sampling unit (correct model). Study findings generally showed that the two models were comparable with respect to desirable statistical properties under a majority of the replicated conditions-more measurement error in ability estimation is found when the intra-class correlation is high and the item pool is small. In practice this is the exception rather than the norm. However, it was found that the empirical Bayes estimates were dependent upon the first-stage sampling unit raising the issue of equity and fairness in educational decision making. A real-world complex survey design with binary outcome data was also fit with both models. Analysis of the data supported the simulation design results which lead to the conclusion that modeling binary Rasch data may resort to a policy tradeoff between desirable statistical properties and measurement validity. / Ph. D.

Dans le contexte latino-américain, façonné par une urbanisation rapide, de fortes inégalités et une faiblesse des institutions de placement de la main d’oeuvre, les problématiques liées à la structure du marché du travail et à l’effet des dispositifs d’intermédiation relationnelle sur les performances dans l’emploi apparaissent fondamentales. Cette thèse se propose de décrire la structure du marché du travail et d’analyser les effets des réseaux de relations sur la qualité de l’emploi et les performances des actifs occupés à Bogota (Colombie). Elle adopte une démarche de recherche pluridisciplinaire reposant sur un cadre d’analyse socioéconomique et institutionnaliste. Dans une première partie, une réflexion théorique et analytique est tout d’abord conduite autour de la notion de qualité de l’emploi envisagée comme un nouvel indicateur de performance. Au prisme de la théorie de la segmentation du marché du travail, l’analyse de ce nouvel indicateur permet d’envisager les logiques et les effets différentiés du recours aux relations sociales. À partir de données quantitatives (enquête ménage geih de 2013) et qualitatives (entretiens collectifs), l’analyse exploratoire multidimensionnelle, économétrique et compréhensive permet de vérifier que : (i.) la qualité de l’emploi traduit une structure fortement polarisée du marché du travail à Bogota, (ii.) l’usage des relations est associé différemment à la qualité de l’emploi des travailleurs en fonction de leur segment, des réseaux de nécessité (segment vulnérable) s’opposant à des réseaux d’opportunité (segment protégé). Dans une seconde partie, s’appuyant sur les théories de l’encastrement et de la sociologie des réseaux, la thèse se propose d’explorer plus précisément les effets des différentes dimensions, configurations et mécanismes de réseau de relations personnelles sur les performances dans l’emploi. À partir d’un système spécifique d’enquêtes mixtes déployé à Bogota entre 2016 et 2018 des données originales de réseaux égocentrés ont été collectées. Les analyses statistiques multidimensionnelles et économétriques ainsi que l’analyse des narrations quantifiées mettent en évidence que : (i.) la combinaison entre un réseau potentiel étendu et un réseau actif cohésif augmente le temps de recherche mais aussi la probabilité de trouver un emploi plus satisfaisant, (ii.) la force des liens apparaît contextualisée et est corrélée négativement avec le revenu et positivement avec l’évolution de ce dernier entre deux emplois, (iii.) au cours des trajectoires professionnelles des acteurs, les ressources nécessaires et les relations permettant d’y accéder se différencient nettement en fonction du type de changement d’emploi (incrémental ou radical). / In the Latin American context, shaped by rapid urbanization, high inequalities and the weakness of labor institutions, issues related to the structure of the labor market and the effect of relational intermediation on job performance appear fundamental. This thesis aims to describe the structure of employment and analyzes the effects of social networks on the quality of employment and the performance of workers in Bogota’s labor market (Colombia). This work adopts a multidisciplinary research approach based on a socioeconomic and institutionalist framework. In the first part, a theoretical and analytical reflection is conducted through the notion of quality of employment, to overcome the classical typologies commonly used in developing countries. From this perspective, quality of employment can be seen as a new performance indicator grasped through the prism of the labor market segmentation theory, making possible to consider the rationales and the differential effects produced by the use of social networks. Subsequently, based on quantitative data from the household survey (geih, 2013) supplemented by information collected through focus groups, the multidimensional, econometric and comprehensive exploratory analysis allows to empirically verify that : (i.) quality of employment reflects a strongly polarized structure of the labor market in Bogota, (ii.) the use of social networks is associated differently with the quality of employment of workers according to their segment ; opposing necessity networks (for the vulnerable segment) and opportunity networks (for the protected segment). Based on the theories of the embeddedness and the sociology of networks, the second part of this thesis proposes to explore the dimensions, configurations and mechanisms of different types of social networks to get a job. Using original data on egocentric networks collected from a specific mixed survey system deployed in Bogota between 2016 and 2018, the empirical results from multidimensional and econometric analyzes and, the application of quantified narratives method demonstrate that : (i.) the combination of an extended potential network and a cohesive active network increases the search time but also the probability of finding a satisfactory job, (ii.) the strength of ties appears contextualized and negatively correlated with income but positively with its evolution between the last and the current job, (iii.) during the actors’ labor market trajectories, the necessary resources for changing job and the relationships to access them are clearly differentiated by the type of evolution (incremental or radical).