La présente thèse s'intéresse à modéliser la causalité dans le cadre de la théorie des fonctions de croyance. Dans un premier temps, nous avons analysé et révisé les fondements théoriques des différents modèles associationnels crédibilistes existants. Nous avons proposé une structure graphique qui sert de base pour le réseau causal crédibiliste. Dans ce dernier, l'incertitude au niveau des nœuds est donnée en termes de distributions de masses conditionnelles. Comme la notion d'intervention est d'une importance capitale pour une analyse causale efficiente, nous avons introduit la contrepartie de l'opérateur \do" pour la représentation des interventions dans les réseaux causaux crédibilistes. Le modèle proposé permet ainsi de calculer l'effet simultané des observations et des interventions. Il est admis que considérer l'intervention comme une action qui réussit toujours à mettre sa cible à une valeur précise en la rendant complètement indépendante de ses causes originales est une condition rarement réalisée dans les applications réelles. De ce fait, nous avons examiné le traitement des interventions dont l'occurrence et/ou les conséquences sont imparfaites. Dans la dernière partie de la thèse, nous avons proposé un modèle d'attribution de causalité permettant d'interpréter les relations d'influence qui existent entre les différents attributs du système à savoir la causalité, la facilitation ou encore la justification en présence de données observationnelles et interventionnelles. Motivé par le fait que les preneurs de décision ne sont pas uniquement intéressés par l'attribution de la causalité, ce modèle permet de définir différentes forces d'une cause. / This thesis focuses on the modeling of causality under the belief function framework. We have first analyzed and revised the theoretical foundations of existing associational belief models. Then, we have proposed a graphical structure that serves as a basis for the causal belief network. In this latter, uncertainty at the nodes is given in terms of conditional mass distributions. Since intervention is a crucial concept for an efficient causal analysis, we have introduced a counterpart of the do operator as a tool to represent interventions on causal belief networks. The proposed model allows computing the simultaneous effect of observations and interventions. It is assumed that considering an intervention as an action that always succeeds to force its target variable to have a precise value, by making it completely independent of its original causes, is a condition rarely achieved in real-world applications. Therefore, we have examined the treatment of interventions whose occurrence is imperfect and/or have imperfect consequences. In the last part of the thesis, we have proposed a model for causality ascription to interpret influential relationships between different attributes of the system namely causality, facilitation or justification in the presence observational and interventional data. Since decision makers are not only interested in ascribing causes, this model allows to define different strengths of a cause.

In this thesis we examine the effect of foreign exchange interventions in small open economy, focusing on the Czech experience. In the first part we model volatility development before and after the intervention using GARCH model. In the second part we estimate relationship between macroeconomical variables using vector autoregressive model. In this part we estimate impulse response function of exchange rate and inflation. In second part of VAR modeling we provide counterfactual analysis, which compare actual development of variables with alternative scenario in which the interventions would not happen . Our results suggest that the interventions is associated with few months delayed decrease in volatility. Base on scenario analysis the interventions increased inflation by approximately 1.5 % and without the intervention the economy would in deflation around -1 % nowadays. KEYWORDS: Vector autoregression, Volatility modelling, Monetary policy, Intervention Author's e-mail: jantimko16@gmail.com Supervisor's e-mail: tomas.holub@cnb.cz

With the collapse of the Soviet Union, humanitarian intervention became an important pillar in the emerging new world order. From 1989 to 1995, 96 violent civil confrontations have occurred, but 91 of them did not result in humanitarian interventions. Here comes the question: Why? Why there were interventions in Iraq, Bosnia, and Kosovo and not in Rwanda, the Sudan, and Tajikistan? These are the main questions that the following study aims to answer. Particularly, the issue of American humanitarian intervention is scrutinized. The casual factors of interventions are examined to explain the selectivity of American Humanitarianism. Furthermore, a theory building is initiated to outline a model of variables which will allow to explain the combination of which casual factors leads to which form of intervention or non-intervention.

Magister Public Health - MPH / Adolescents living with HIV (ALHIV) are notably underserved by national HIV programmes globally due to their unique needs. Of particular concern is limited access to and availability of adolescent-friendly ART services, which contributes to poor ART adherence and retention in care in many sub-Saharan African countries. Poor adherence in adolescents has been associated with medicine side effects, pill fatigue, non-disclosure of status to the child, inadequate information on HIV, caregiver-child communication, caregiver’s health beliefs and stigma, and lack of knowledge on the rationale of taking medicines. Several interventions have been developed to improve ART adherence and retention in care amongst ALHIV through peer groups and psychosocial support. The Teen Club intervention was introduced in 2010 at Intermediate Hospital Katutura Paediatric ART clinic in Windhoek to improve ART adherence and retention in care amongst ALHIV by providing psychosocial support in a group environment. However, to date no formal evaluation of the effectiveness of the Teen Club intervention in Namibia has been conducted. The aim of the study was to compare the effects of the Teen Club intervention against standard care on treatment outcomes for ART (i.e. adherence, retention in care and viral suppression) in adolescents at Intermediate Hospital Katutura Paediatric ART clinic in Namibia. Methods: A retrospective cohort analysis of HIV positive adolescents aged 10-19 years, who were accessing ART between 1 July 2015 and 30 June 2017 was conducted. Patient data was extracted from the electronic Patient Monitoring System (ePMS), individual Patient Care Booklets and the teen club attendance register. Adherence to ART was measured through pill counts; and retention by kept clinic visits. Viral load results were assessed to measure levels of viral suppression. Adolescents with viral loads ≥ 1000 copies/ml were classified as not virally suppressed whilst those with viral loads <1000 are virally suppressed (with those <40 fully suppressed). Results: The total sample was 385 participants; with 78 of them in the Teen Club (exposed) and 307 adolescents in standard care (unexposed).

Magister Artium (Child and Family Studies) - MA(CFS) / Fathers have a valuable, nurturing role to play in the lives of their children. Although a father’s availability and time with his child is important, the quality of a father’s involvement with his child is however, the strong predictor of child well-being. Therefore, Fatherhood intervention programmes desire to address fundamental issues that prevent men from succeeding in their fathering role. An understanding and evidence of the influence of fatherhood interventions on the role of the father in the family, is thus required. The aim of the study was to explore the perceptions, experiences and challenges of families after the father’s exposure to a fatherhood intervention programme. A qualitative approach with semi-structured interviews was utilized. Thematic analysis had been used to analyse the data and five themes emerged from the study. They were constructions of fatherhood, experiences and perceptions of the fatherhood intervention programme, father’s experiences of the father-child relationship after exposure to the programme, spouse/partner perceptions and experiences of their relationships with their spouse after exposure to the programme, and the facilitator’s experiences of engaging fathers. The overall outcomes were that families perceived and experienced the fathers to be more involved, responsible and sharing in parenthood after exposure to the fatherhood intervention programme. This led to an increase of father involvement with children - childcare activities, schoolwork and well-being of the children; their spouse/partner - an improvement in the quality of the relationship between the fathers and their spouses/partners and an increase involvement in the home.

Student Number : 0206926T - PhD thesis - School of Geography, Archaeology and Environmental Studies - Faculty of Science / The overarching objective of this thesis is to determine causes of food insecurity in southern Africa, and how it can best be addressed. This objective is addressed through a number of research questions and methods at three geographic scales: the regional, through a technique of meta-analysis which is used to synthesise 49 local-level household economy case studies; the provincial, through a Delphi panel of practitioner experience; and the local, using multiple research techniques, including participatory methods. An extremely diverse range of factors contributing to food-insecurity are found at all three scales, indicating that community- and household-specific dynamics give rise to forms of food insecurity. Two common processes, however, are argued to be common across all the casestudy communities in the regional-scale research. These are the closely related processes of cycles of intensifying vulnerability associated with livelihood ‘trade-offs’, and of communitylevel social capital changing into forms that undermine resilience to food insecurity - such as the decline in two-parent families. A further probing of social capital at the local level suggests that while social capital takes multiple forms, and further remains in many respects a problematic concept, it nevertheless provides a valuable lens through which powerful social dynamics might be examined in developing responses to food insecurity. Policy makers and change agents should carefully consider their role in building community social-capital that might enhance the ability of vulnerable communities to overcome livelihood constraints and adapt to the tremendous challenges posed by changing economic environments in southern Africa. Drawing on the research at all scales, a framework is provided that calls for a reconceptualisation of food-security interventions to focus on intervention processes, applicable at all scales and in all contexts across the region. The development of social capital, participation, co-ordination and learning interactions are explored as central elements in these processes. The framework asks for closer attention to both the appropriate mechanisms (such as policy) necessary to effect change, and the human dimensions that give these mechanisms agency. The findings of the thesis represent an additional shift in understanding food security to acknowledge that the value of a political economic interpretation of food security is limited independent of an understanding of the cross-scale social networks and relational interactions that ultimately configure and reconfigure it.

Abstract South Africa has one of the largest population of adults living with HIV. About 30 to 60% of people living with HIV have been found to have HAND which can affect treatment with HAART which requires 95% compliance rate in order to be effective. Interventions that seek to alleviate the cognitive deterioration that is associated with HIV can include Internet based rehabilitation programmes. However, internet based interventions are plagued by poor adherence and attrition rates. The aim of the study was to describe the challenges and facilitating factors in the process of implementing the CogMed™ Working Memory Training Programme at a public HIV clinic for adults living with HIV. The study used a qualitative method with an ethnographic approach. Data was collected through CogMed™ administrator console, observations, interactions with gatekeepers, interactions and interviews with participants and interviews healthcare workers. The factors involved in the implementation process were categorised into the four “Stages of Use”. In the Recruitment/Consideration Stage perception of need, identifying and defining an ideal user were the main themes. Factors affecting the Initiation of Use Stage were sense of obligation, time to commit and access to suitable training environment. Utilisation of Service was influenced by ease of drop out, ease of use, perceived cost versus the perceived benefits of participating. Predisposing, enabling, and needs factors that affected adherence and participation were explored. Finally the Outcomes Stage covered the experiences and perceptions of using the rehabilitation tool. Limitations of the study were also discussed. Keywords: CogMed, Working Memory, HIV, HAND, Internet based interventions, Stages of Use

Thesis advisor: Mary E. Walsh / The out-of-school factors that low-income children face can impact their wellbeing and ability to learn (Rothstein, 2010), leading to low academic performance, and, in turn, high levels of stress among their teachers. One of the numerous potential strategies that exist to address this problem is the implementation of systemic student support interventions, which are hypothesized to support teachers in addition to students (Ball & Anderson-Butcher, 2014). Using the City Connects intervention as an example of a comprehensive, systemic student support intervention that has demonstrated positive effects for students, this dissertation study examines the impact of this same intervention on teachers. The study used data obtained from annual surveys administered to all teachers in public schools (across several districts) where the City Connects intervention was implemented. The sample consisted of 656 responses from teachers in Boston Public Schools that were participating in the City Connects intervention during three designated school years (2012-13, 2013-14, and 2015-16). The study implemented a mixed-methods approach that allowed for an in-depth analysis of teachers’ perceptions of support from the City Connects intervention through both quantitative and qualitative data sources. Following exploratory and confirmatory factor analyses of the survey instrument itself, survey data was analyzed using the Validating Quantitative Data Model (Creswell & Plano Clark, 2007), in which quantitative findings were confirmed and expanded upon through analysis of data from a small number of open-ended survey questions. The results of this dissertation study support the continued use of systemic student support interventions in schools, as data confirm that teachers in schools with City Connects report being supported by the intervention. These results hold for veteran teachers who have been teaching for over sixteen years and have participated in the City Connects intervention for over ten years. Ultimately, the findings of this dissertation study suggest that, in addition to leading to positive changes in students’ developmental trajectories, student support interventions can improve the experiences of other members of the school community. / Thesis (PhD) — Boston College, 2018. / Submitted to: Boston College. Lynch School of Education. / Discipline: Counseling, Developmental and Educational Psychology.

Thesis advisor: Mary E. Walsh / Developmental-contextual theory asserts that the contexts in which children develop have the potential to foster or inhibit healthy development (Lerner, 1984; 1995). Given the potential for developmental contexts to promote positive development for at-risk children, systemic student support interventions have been developed to change school contexts into more supportive environments for healthy child development (e.g., BCCOSS, 2010; Communities in Schools, 2010; Dryfoos, 2003). The current study examined the effects of one such student support program, City Connects (formerly Boston Connects). Previous evaluation research has found that City Connects positively impacts multiple factors of child development (BCCCFCP, 2009; BCCOSS, 2010). This study expanded upon the evaluation research to investigate the relationship between the City Connects intervention and specific domains of social development: friendship, bullying, peer victimization, relationship with teacher, and school belonging. Positive social development has been found to foster resilience and promote positive child development in other domains. Thus, this study also examined these domains of social development as the mechanisms through which City Connects is related to student report card grades in math, reading, behavior, and work habits. Within the current study sample of 3rd through 5th grade students in 2007, significant direct relationships between City Connects and domains of social development were not found. Since City Connects did not significantly predict improvements in social development, the mediating hypothesis was not supported. However, follow-up analyses revealed indirect relationships between City Connects and domains of social development, which were mediated by report card grades in reading and work habits. Reading grades significantly mediated the relationships between City Connects and school belonging, peer victimization, and bullying behavior. Work habit grades approached significance as a mediator of the relationships between City Connects and school belonging, relationship with teacher, bullying behavior and peer victimization. The current study underscores the complexity of developmental pathways, and the need for complex, multifaceted student support interventions to help support positive child development for low-income, urban children. / Thesis (PhD) — Boston College, 2011. / Submitted to: Boston College. Lynch School of Education. / Discipline: Counseling, Developmental, and Educational Psychology.

Syftet med studien var att, genom granskning av vetenskapliga artiklar, undersöka olika former av arbetsterapeutiska interventioner för barn med Downs syndrom. Insamlingen av data genomfördes genom en litteratursökning via databaser som fanns att tillgå via Luleå tekniska universitet. Artiklarna kvalitetsgranskades för att avgöra om de var lämpliga att inkludera i studien. Totalt identifierades 18 artiklar som legat till grund för studien. Resultatet visade att det finns arbetsterapeutiska interventioner för barn med Downs syndrom. De identifierade arbetsterapeutiska interventionerna kategoriserades utifrån Fishers (2009) sex interventionsformer: “preparation”, “rote practice/exercise”, “simulated occupation”, “restorative occupation”, “acquistional occupation” och “adaptive occupation”. Dock inkluderades artiklarnas interventioner endast av fyra av Fishers (2009) sex interventionsformer. Vidare undersöktes det om artiklarnas arbetsterapeutiska interventioner skedde på aktivitets-eller funktionsnivå. Resultatet av litteraturöversikten visade på att den större andelen av de arbetsterapeutiska interventionerna för barn med Downs syndrom sker på funktionsnivå, det vill säga utifrån de interventioner som Fisher (2009) beskriver som: “preparation”, “rote practice/exercise” och “simulate / The purpose of the study was to investigate available types of occupational therapeutic interventions for children with Down's syndrome by reviewing scientific articles. Data collection was carried out through a literature search through databases that were available through Luleå University of Technology. The articles were quality-reviewed to determine if they were suitable for inclusion in the study. In total, 18 articles were identified as the basis for the study. The results showed that there are occupational therapy interventions for children with Down's syndrome. The identified occupational therapeutic interventions were categorized based on Fisher's (2009) six forms of intervention: "preparation", "rote practice / exercise", "simulated occupation", "restorative occupation", "acquistional occupation" and "adaptive occupation". However, the interventions of the articles were included only by four of Fisher's (2009) six intervention forms. Furthermore, it was examined whether the articles' occupational therapeutic interventions took place at the level of activity or function. The result of the literature review showed that the larger proportion of the occupational therapeutic interventions for children with Down syndrome occurs at the functional level, that is, based on the interventions that Fisher (2009) describes as: "preparation", "rote practice / exercise" and "simulated occupation”.