[pt] A seca é um dos desastres naturais mais críticos que tem efeitos devastadores sobre habitats naturais, ecossistemas e muitos setores econômicos e sociais. Devido a esses graves impactos dos eventos de seca, muitos estudos estão focados no monitoramento, previsão e análise de risco de secas para auxiliar os planos de preparação e medidas de mitigação. Esta tese propõe um sistema de monitoramento e alerta de secas na região do semiárido do Brasil, chamado Drought Risk Assessment Interface (DRAI), que se baseia em um índice composto de risco de seca. O índice de risco possui duas componentes: ameaça e vulnerabilidade. A ameaça considera indicadores meteorológicos, enquanto a vulnerabilidade considera variáveis sociais. Com base na opinião de especialistas de vários países do mundo, com mais de 10 anos de experiência na área, foi definido o peso de cada um desses indicadores usando o processo de hierarquia analítica (AHP - Analytical Hierarchy Process). Os resultados foram comparados com outros índices de seca com o intuito de validar o índice proposto. Em seguida, foram levantados os principais sistemas de monitoramento e alertas em nível nacional e internacional e, então, foi proposto um padrão para geração de alertas no DRAI. Os alertas foram associados à sete medidas de mitigação de risco de seca validadas por técnicos locais. O DRAI tem como usuário final, além de outros pesquisadores, as Defesas Civis que poderão atuar diretamente nas ações de mitigação dos riscos. Como pesquisas futuras, sugere-se a automatização da coleta dos dados que compõem o índice de ameaça bem como a aplicação do estudo para todo o território brasileiro. / [en] Drought is one of the most critical natural disasters that have devastating effects on natural habitats, ecosystems and many economic and social sectors. Due to these severe impacts of drought events, many studies are focused on monitoring, forecasting and analyzing drought risk, to help with drought preparedness plans and mitigation measures. This study presents a drought early warning system in the semiarid region of Brazil, called the Drought Risk Assessment Interface (DRAI), which is based on a composite index of meteorological drought risk. The risk index has two components: hazard and vulnerability. The hazard considers meteorological indicators while the vulnerability considers social variables. Based on the opinion of experts from several countries in the world, with more than ten years of experience in the field, we define the weight of each of these indicators using the analytical hierarchy process (AHP). Then, the main early warning systems at national and international level were raised and then, a standard for generating warnings in the DRAI was proposed. The warnings were associated with seven drought risk mitigation measures validated by local technicians. DRAI has as its end user, in addition to other researchers, Civil Defenses that can act directly in risk mitigation actions. Finally, the system and its main features are presented. As future research, we suggest automating the collection of data that make up the hazard index as well as applying the study to the entire Brazilian territory.

Inequalities in levels of health between regions within a country are frequently regarded as a problem. Zimbabwe is characterised by poor and unequal conditions of health (both the state of people’s health and health services). The health system of the country shows severe spatial inequalities that are manifested at provincial, district and even local levels. This research therefore examines and analyses the spatial inequalities and temporal variation of health conditions in Zimbabwe. Composite indices were used to determine the people’s state of health in Zimbabwe. Administrative districts were ranked according to the level of people’s state of health. Cluster analysis was also performed to demarcate administrative districts according the level of health service provision. Districts with minimum difference were demarcated in a single cluster. Clusters were delineated using data on patterns of diseases and health and such clusters were used to demarcate the country’s spatial health system according to the Adapted Epidemiological Transition Model. This was used to evaluate the applicability of the model to Zimbabwe. It emerged from the research that generally the country’s health conditions are poor and the health system is characterised by severe spatial inequalities. Some districts are experiencing poor health service provision and serious health challenges and are still in the age of pestilence and famine but others have good health service provision as well as highly developed health conditions and are in the age degenerative diseases of the epidemiological transition model. It further emerged that the country’s health has been evolving with signs of improvement since the 1990s. Recommendations were made regarding possible adjustment to previous strategies and policies used in Zimbabwe, for the development of the health system of the country. New strategies were also recommended for the improvement of the health system of the country. Some proposals are made for further research on the spatial development of health in the country. / Geography / D. Litt et. Phil. (Geography)

本研究以2003年到2012年國內上市上櫃公司作為研究對象，探討公司治理與公司績效之關聯性，以及家族與非家族企業公司治理對於公司績效之差異性。公司治理衡量方式是參考葉銀華、李存修與柯承恩（2002），設定一個公司治理綜合指標，綜合指標下有五個構面，為董監事組成、股東權益、管理型態、關係人與介入股市構面，各構面下有其所屬變數，將變數評分後，彙總於該變數所屬構面，最後再依照各構面之重要程度加權過後，即形成公司治理綜合指標。實證結果如下： 一、以全體企業、家族企業和非家族企業為樣本群，皆顯示當公司治理越好時，公司績效越好。 二、在資產報酬率與每股盈餘應變數下，家族企業之公司治理綜合指標對公司績效正向影響程度小於非家族企業。 三、以公司治理綜合指標下之五個構面，分別去探討家族企業與非家族企業對於公司績效之差異。董監事組成構面下之家族企業對公司績效正向影響程度大於非家族企業；股東權益、管理型態、關係人與介入股市構面之各個實證結果，皆是家族企業對公司績效正向影響程度小於非家族企業。 / This study investigates the relation between corporate governance and performance and the difference in that relation between family and non-family ownership using a sample of publicly-traded Taiwanese companies during 2003-2012. Based on Ye, Lee, and Ke (2002, in Chinese), a composite index of corporate governance is set up from five constructs including director and supervisor, shareholders' equity, management style, overinvestment and related party. The empirical results are as follows: 1.The better the corporate governance, the better the corporate performance. This finding is true for each of the three samples including family ownership, non-family ownership and both of them together. 2.Taking return on assets (ROA) or earnings per share (EPS) as corporate performance measure, the effect of corporate governance on performance is stronger for family ownership than for non-family ownership. 3.Except for director and supervisor construct, all other four constructs of corporate governance including shareholders' equity construct, management style construct, overinvestment construct and related party construct have stronger effect on corporate performance for non-family ownership than for family ownership.

Inequalities in levels of health between regions within a country are frequently regarded as a problem. Zimbabwe is characterised by poor and unequal conditions of health (both the state of people’s health and health services). The health system of the country shows severe spatial inequalities that are manifested at provincial, district and even local levels. This research therefore examines and analyses the spatial inequalities and temporal variation of health conditions in Zimbabwe. Composite indices were used to determine the people’s state of health in Zimbabwe. Administrative districts were ranked according to the level of people’s state of health. Cluster analysis was also performed to demarcate administrative districts according the level of health service provision. Districts with minimum difference were demarcated in a single cluster. Clusters were delineated using data on patterns of diseases and health and such clusters were used to demarcate the country’s spatial health system according to the Adapted Epidemiological Transition Model. This was used to evaluate the applicability of the model to Zimbabwe. It emerged from the research that generally the country’s health conditions are poor and the health system is characterised by severe spatial inequalities. Some districts are experiencing poor health service provision and serious health challenges and are still in the age of pestilence and famine but others have good health service provision as well as highly developed health conditions and are in the age degenerative diseases of the epidemiological transition model. It further emerged that the country’s health has been evolving with signs of improvement since the 1990s. Recommendations were made regarding possible adjustment to previous strategies and policies used in Zimbabwe, for the development of the health system of the country. New strategies were also recommended for the improvement of the health system of the country. Some proposals are made for further research on the spatial development of health in the country. / Geography / D. Litt et. Phil. (Geography)

Gleichzeitig mit der weit verbreiteten Verwendung von traditionellen einkommens- oder konsumbasierten Maßnahmen zur Messung menschlicher Armut und Entwicklung gibt es seit den späten 70er Jahren zunehmendes Interesse an der Ökonomie des Glücks. Ebenso gibt es einen breiten Literaturbereich, der Indizes definiert und diskutiert hat, die „functionings" auf der Grundlage des „Capabilities Approach“ von Sen (1985) auf sich vereinen, welche auf einer Vielzahl von ideologischen Urteilen und Zielen zur Bestimmung des objektiven Wohlbefindens (1984), S. 187) beruhen. Beide Ansätze sind ähnlich in ihrer Prämisse, dass Einkommen oft eine unzureichende Determinante des Wohlbefindens ist – ein latenter Begriff, dass besser mit anderen, breiteren Definitionen- subjektiv oder objektiviert- erfasst wird. Beide Konzepte des Wohlbefindens wurden im Hinblick auf ihr Verhältnis zum Einkommen untersucht, und es wurde ein klarer Unterschied zwischen Einkommen und diesen beiden Maßnahmen festgestellt. Angesichts der relativen Neuheit und Komplexität beider Ansätze sind sie jedoch bislang selten in wissenschaftlichen Arbeiten zusammengebracht worden. Den dritten Aufsatz in dieser Arbeit ist ein Versuch, diese beiden Ansätze zu kombinieren und damit diese Lücke in der Literatur zu erfüllen. Subjektives Wohlbefinden wird mit verfügbaren Daten, die Zufriedenheit messen, festelegt, während das objektive Wohlbefinden durch einen Multidimensional Poverty Index (MPI) operationalisiert wird (Alkire & Santos, 2010). Der MPI ist einer der neuesten Versuche zur Messung des menschlichen Wohlbefindens im Rahmen des Capabilities Approach. Die Entwicklungspolitik erkennt ebenfalls graduell diese Maße als eine genauere Beschreibung des Wohlbefindens an, oder betrachtet diese zumindest als eine sinnvolle Ergänzung zu metrisch-monetären Maßen. Angesichts der vielen nationalen und internationalen Programme, die eine umfassende Verbesserung des menschlichen Wohlbefindens zum Ziel haben, gibt es erstaunlich wenige Arbeiten, die dazu beitragen können, die Auswirkungen eines bestimmten Programms auf das allgemeine Wohlbefinden und nicht nur auf eine bestimmte Dimension zu quantifizieren und zu bewerten. Der zweite Aufsatz in dieser Arbeit beschäftigt sich kritisch mit diesem Ansatz und betrachtet dabei den besonderen Fall Südafrikas. Aufgrund der steigenden Beliebtheit der mehrdimensionalen Armutsmaße besteht ein zunehmender Bedarf an einer Überprüfung ihrer grundlegenden Eigenschaft, ebendies zu erreichen. Eine wachsende Zahl von Forschern hat sich mit den Problemen beschäftigt, die ein zusammengesetztes Maß wie der MPI mit sich bringen kann und dessen Fähigkeit zur Messung multidimensionalen Wohlbefindens beeinträchtigen kann. Ein bestimmter Aspekt ist hierbei die Gewichtung der einzelnen Dimensionen und Indikatoren, um Armut über verschiedene Regionen hinweg zu definieren. Diese Dissertation schafft in Aufsatz eins eine Brücke zwischen den statistischen Methoden und den optimalen Gewichtungsschemata, die speziell zur Messung des multidimensionalen Wohlbefindens in verschiedenen Ländern genutzt werden können.