The overall aim of this thesis is to contribute to an understanding of the pathways through which structural and intermediary determinants influence child health in Colombia and how they operate through the context where children live in a more comprehensive way than has been determined to date. Colombia has made significant progress in child health in the last few decades and it is currently on track to meet the Millennium Development Goals (MDG). Nearly 90% of the goals on global malnutrition, infant mortality rate and under-five mortality rate have been achieved. However, despite the progress, national averages remain masking huge territorial disparities. While some regions present figures similar to those of a developed country, others report indicators similar to those of a very poor African country. Some municipalities, for example, record no stunted children, whereas in others, the prevalence of chronic malnutrition is greater than 50%. In this context, empirical research that enhances our understanding of socioeconomic determinants of child health and guide policy-making is crucial in order to reduce place-based health inequities in Colombia. This thesis is based on three research articles. The data used in these studies are drawn from the 2010 Colombian Demographic and Health Survey (DHS). The DHS is a large survey programme designed to collect high-quality nationally representative data on population, health and nutrition for developing countries. The DHS are widely recognised as the most important source of information for the analysis of health inequities in the developing world. Firstly, intermediary determinants of early childhood health have been analysed through the construction of a composite index (Chapter 3). The intermediary determinants are the most immediate mechanisms through which socioeconomic position influence child health inequities. Therefore, their identification should contribute to the drafting of intervention policies at this level, given the importance of these factors in programmes aimed at improving maternal and child care. The index allowed us to identify key immediate determinants of child health and their relative importance among Colombian departments (administrative subdivisions). The index was constructed using a more sophisticated methodological approach than that commonly used in the literature, termed polychoric principal component analysis (PCA). A hierarchical cluster analysis was also carried out in order to identify how departments cluster based on the health of their children rather than their geographic proximity. The results showed that the largest differences in intermediary determinants of child health are associated with health care before and during delivery. Furthermore, the departments that perform relatively better in the most immediate determinants of child health are located in the centre of the country. In contrast, those departments that perform worse are located in the peripheral region. This region has a per capita gross domestic product (GDP) well below the national average, little state presence, a hostile environment and a large proportion of the ethnic minorities. Our index provides very useful information in terms of public policy since it facilitates measuring, visualizing and monitoring of child health indicators, and may, therefore, help identify potential intervention strategies for improving the well-being of Colombian children. Secondly, this thesis examines the effect of individual, family and community socioeconomic conditions on different indices representing intermediary determinants of child health, using a coherent conceptual framework (Chapter 4). Using a weighted multilevel approach, the results indicate that whilst community socioeconomic context can exert a greater influence on factors linked directly to health, in the case of psychosocial factors and parent’s behaviours, the family context can be more important. In addition, the results indicate that a significant percentage of the variability in the overall index of intermediary determinants of child health is explained by the community context, even after controlling for individual, family and community characteristics. This study provides evidence that community socioeconomic context is a key component for improving child health in Colombia. However, the role played by context may vary according to the category of the intermediary determinants of child health analysed, highlighting the importance of distinguishing between community and family intervention programmes. Thirdly, the influence of education of other women in the community and family socioeconomic characteristics on child nutrition outcomes, as well as their interactions, was investigated (Chapter 5). The contextual effects of education on child health were studied using weighted multilevel models. This study takes into account important methodological issues such as sample weights and second level endogeneity in multilevel modelling, which have not been addressed in the empirical literature and can lead to biases in the estimates. Different ways through which community education can substitute for the effect of family characteristics on child nutrition were found, suggesting that child care programmes should focus not only on individuals but should also target the broader context of communities. In particular, those communities with less educated mothers and with low female autonomy are those that could benefit more from intervention policies that focus on encouraging female education. To conclude, previous studies on the social determinants of child health in Colombia are limited. Most of them have covered the issue from the perspective of the individual and little attention has been paid to the effect of context. In this vein, understanding the structural and intermediary determinants of child health inequities, as well as the role played by community socioeconomic context, is essential for the design, monitoring and tracking of public child care policies in Colombia. / El objetivo de esta tesis es contribuir a un entendimiento de los mecanismos a través de los cuales los determinantes estructurales e intermediarios de la salud influencian la salud de los niños en Colombia y como estos operan a través del contexto donde los niños viven, de una manera más completa que la que se ha hecho hasta ahora. Esta tesis se base en tres artículos de investigación. Los datos usados en estos estudios son de la “Demographic and Health Survey” (DHS) para Colombia en el año 2010. En primer lugar, los determinantes intermediarios de salud infantil se analizan a través de un indicador compuesto. El indicador permite identificar determinantes intermediarios claves de salud infantil y su importancia relativa entre departamentos colombianos. En segundo lugar, esta tesis examina el efecto de las condiciones socioeconómicas de la comunidad, la familia y los individuos sobre diferentes índices que representan determinantes intermediarios de salud infantil. Por último, se analiza la influencia de la educación de otras mujeres en la comunidad y de características socioeconómicas de las familias sobre la nutrición infantil, así como sus interacciones. Se utilizan diferentes enfoques metodológicos. Para la construcción de los indicadores compuestos se utiliza un análisis de componentes principales usando correlaciones policóricas, con el fin de analizar cómo se agrupan los departamentos de acuerdo a la salud de sus niños se hace un análisis de cluster jerarquico. Finalmente, con el fin de medir efectos contextuales se utilizan modelos multinivel de regresión lineal ponderados. Los resultados de esta tesis demuestran que el contexto de la comunidad es un componente clave en la determinación de determinantes intermediarios de salud infantil, así como de indicadores de nutrición de los niños, tanto en el mediano como en el largo plazo en Colombia. Por lo tanto, es necesario que los gobiernos municipales y departamentales involucren no sólo a los individuos, sino que también tengan en cuenta el contexto de las comunidades para el desarrollo, ejecución, monitoreo y evaluación de las políticas de cuidado infantil.

El objetivo de esta tesis es predecir el rendimiento de los estudiantes de doctorado en la Universidad de Girona según características personales (background), actitudinales y de redes sociales de los estudiantes. La población estudiada son estudiantes de tercer y cuarto curso de doctorado y sus directores de tesis doctoral. Para obtener los datos se ha diseño un cuestionario web especificando sus ventajas y teniendo en cuenta algunos problemas tradicionales de no cobertura o no respuesta. El cuestionario web se hizo debido a la complejidad que comportan de las preguntas de red social. El cuestionario electrónico permite, mediante una serie de instrucciones, reducir el tiempo para responder y hacerlo menos cargado. Este cuestionario web, además es auto administrado, lo cual nos permite, según la literatura, unas respuestas mas honestas que cuestionario con encuestador.Se analiza la calidad de las preguntas de red social en cuestionario web para datos egocéntricos. Para eso se calcula la fiabilidad y la validez de este tipo de preguntas, por primera vez a través del modelo Multirasgo Multimétodo (Multitrait Multimethod). Al ser datos egocéntricos, se pueden considerar jerárquicos, y por primera vez se una un modelo Multirasgo Multimétodo Multinivel (multilevel Multitrait Multimethod). Las la fiabilidad y validez se pueden obtener a nivel individual (within group component) o a nivel de grupo (between group component) y se usan para llevar a cabo un meta-análisis con otras universidades europeas para analizar ciertas características de diseño del cuestionario. Estas características analizan si para preguntas de red social hechas en cuestionarios web son más fiables y validas hechas "by questions" o "by alters", si son presentes todas las etiquetas de frecuencia para los ítems o solo la del inicio y final, o si es mejor que el diseño del cuestionario esté en con color o blanco y negro.También se analiza la calidad de la red social en conjunto, en este caso específico son los grupos de investigación de la universidad. Se tratan los problemas de los datos ausentes en las redes completas. Se propone una nueva alternativa a la solución típica de la red egocéntrica o los respondientes proxies. Esta nueva alternativa la hemos nombrado "Nosduocentered Network" (red Nosduocentrada), se basa en dos actores centrales en una red. Estimando modelos de regresión, esta "Nosduocentered network" tiene mas poder predictivo para el rendimiento de los estudiantes de doctorado que la red egocéntrica.Además se corrigen las correlaciones de las variables actitudinales por atenuación debido al pequeño tamaño muestral. Finalmente, se hacen regresiones de los tres tipos de variables (background, actitudinales y de red social) y luego se combinan para analizar cual para predice mejor el rendimiento (según publicaciones académicas) de los estudiantes de doctorado. Los resultados nos llevan a predecir el rendimiento académico de los estudiantes de doctorado depende de variables personales (background) i actitudinales. Asimismo, se comparan los resultados obtenidos con otros estudios publicados. / The aim of this Doctoral Thesis is to predict the PhD students' academic performance in the University of Girona from characteristics of their research group understood as a social network and from background and attitudinal characteristics of the PhD student.The data collection was done through web survey, regarding the traditional problems related to coverage and response errors. Web survey was made in order to reduce the complexity of social network questions.The quality of social network questions for questionnaires with egocentered data is analyzed. For this, reliability and validity of this kind of questions are computed, for first time using a Multilevel Multitrait-Multimethod approach. Data are egocentered and thus hierarchy can be used. Reliability and validity can be obtained in an individual level (within group component) or group level (between group component), which are used for carrying out a meta analysis with other European universities. We study if social network questions asked in web questionnaires are more reliable and valid done "by questions" or "by alters", if the frequency labels are present or not and if the best questionnaire design is plane or graphical.Social network as a whole is also studied, in this case are the research groups in the University of Girona. There is a proposal for the missing data problem in complete networks, different from the egocentered network. This new network structure is called "Nosduocentered Network", which is based in two central actors in a network. Regression model estimations explain that this network has more predictive power in order to predict the academic performance for PhD students.Finally, regression models with background, attitudinal and social network variables are done. Then, we combine the regressions in order to specify the best combination which predicts the PhD student's performance (according to academic publications) best. According to the results, the academic performance depends on background and attitudinal variables. Also, a comparison of results from other published studies is done.