Cette thèse se place dans le contexte de la vérification déductive des programmes et a pour objectif de formaliser un certain nombre de concepts qui sont mis en œuvre dans l'outil de vérification Why3.L'idée générale est d'explorer des solutions qu'une approche à base de systèmes de types peut apporter à la vérification. Nous commençons par nous intéresser à la notion du code fantôme, une technique implantée dans de nombreux outils de vérification modernes, qui consiste à donner à des éléments de la spécification les apparences d'un code opérationnel. L'utilisation correcte du code fantôme requiert maintes précautions puisqu'il ne doit jamais interférer avec le reste du code. Le premier chapitre est consacré à une formalisation du code fantôme, en illustrant comment un système de types avec effets en permet une utilisation à la fois correcte et expressive. Puis nous nous intéressons à la vérification des programmes manipulant des pointeurs. En présence d'aliasing, c'est-à-dire lorsque plusieurs pointeurs manipulés dans un programme dénotent une même case mémoire, la spécification et la vérification deviennent non triviales. Plutôt que de nous diriger vers des approches existantes qui abordent le problème d'aliasing dans toute sa complexité, mais sortent du cadre de la logique de Hoare, nous présentons un système de types avec effets et régions singletons qui permet d'effectuer un contrôle statique des alias avant même de générer les obligations de preuve. Bien que ce système de types nous limite à des pointeurs dont l'identité peut être connue statiquement, notre observation est qu'il convient à une grande majorité des programmes que l'on souhaite vérifier. Enfin, nous abordons les questions liées à la vérification de programmes conçus de façon modulaire. Concrètement, nous nous intéressons à une situation où il existe une barrière d'abstraction entre le code de l'utilisateur et celui des bibliothèques dont il dépend. Cela signifie que les bibliothèques fournissent à l'utilisateur une énumération de fonctions et de structures de données manipulées, sans révéler les détails de leur implémentation. Le code de l'utilisateur ne peut alors exploiter ces données qu'à travers un ensemble de fonctions fournies. Dans une telle situation, la vérification peut elle-même être modulaire. Du côté de l'utilisateur, la vérification ne doit alors s'appuyer que sur des invariants de type et des contrats de fonctions exposés par les bibliothèques. Du côté de ces dernières, la vérification doit garantir que la représentation concrète raffine correctement les entités exposées, c'est-à-dire en préservant les invariants de types et les contrats de fonctions. Dans le troisième chapitre nous explorons comment un système de types permettant le contrôle statique des alias peut être adapté à la vérification modulaire et le raffinement des structures de données. / This thesis is conducted in the framework of deductive software verification.is aims to formalize some concepts that are implemented in the verification tool Why3. The main idea is to explore solutions that a type system based approach can bring to deductive verification. First, we focus our attention on the notion of ghost code, a technique that is used in most of modern verification tools and which consists in giving to some parts of specification the appearance of operational code. Using ghost code correctly requires various precautions since the ghost code must never interfere with the operational code. The first chapter presents a type system with effects illustrating how ghost code can be used in a way which is both correct and expressive. The second chapter addresses some questions related to verification of programs with pointers in the presence of aliasing, i.e. when several pointers handled by a program denote a same memory cell. Rather than moving towards to approaches that address the problem in all its complexity to the costs of abandoning the framework of Hoare logic, we present a type system with effects and singleton regions which resolves a liasing issues by performing a static control of aliases even before the proof obligations are generated. Although our system is limited to pointers whose identity must be known statically, we observe that it fits for most of the code we want to verify. Finally, we focus our attention on a situation where there exists an abstraction barrier between the user's code and the one of the libraries which it depends on. That means that libraries provide the user a set of functions and of data structures, without revealing details of their implementation. When programs are developed in a such modular way, verification must be modular it self. It means that the verification of user's code must take into account only function contracts supplied by libraries while the verification of libraries must ensure that their implementations refine correctly the exposed entities. The third chapter extends the system presented in the previous chapter with these concepts of modularity and data refinement.

Les sociologues s’accordent pour constater l’émergence de nouveaux rapports au temps, mais débattent des marges de manœuvre différenciées dont disposent les individus face à ces phénomènes. L’hypothèse que la vitesse de déplacement puisse être un levier est confortée par les études montrant que nos sociétés ont maîtrisé l’espace-temps grâce à l’amélioration des transports d’information et de marchandises. Cependant, peu de travaux empiriques abordent le sujet à l’échelle de l’individu. Cette thèse, qui repose sur une combinaison de méthodes qualitatives et quantitatives, a pour objectif de compléter ces recherches à travers l’étude des modes de transport utilisés et des pratiques spatiales quotidiennes en Île-de-France. Cette approche micro-sociale montre que l’augmentation du nombre et de la durée des activités hors du domicile n’est pas corrélée à une hausse des vitesses, mais à une diminution des distances parcourues pour se rendre à certaines activités. Ceci s’explique par les logiques de localisation des activités des individus. Pour celles auxquelles ils accordent de la valeur, ils maximisent la qualité de la destination, ce qui passe par l’utilisation de modes motorisés rapides. Pour les activités secondaires, ils minimisent leur temps de trajet en les localisant à proximité d’activités structurantes, ce qui induit l’usage de modes plus lents. Comme l’articulation entre ces deux logiques est plus aisée dans Paris qu’en banlieue, l’adoption d’un rythme de vie plus élevé et d’une planification plus souple y est facilitée. Plus que l’accès à la vitesse, ce sont les territoires de résidence qui induisent des disparités dans les marges de manœuvre face au temps / Times studies agree on the fact that our society is currently undergoing changes in terms of pace of life, but debates arise as to the uniformity of such changes within the population and individuals’ scope of action in the face of said phenomena. Given that society as a whole mastered time and space through the rise of transport speed, one is tempted to assume that it is also the case on the individual level. However, few empirical studies have taken it upon themselves to analyze the impact of speed in daily mobility on individuals’ pace of life. Using a combination of statistical analysis and in-depth interviews, this thesis aims at providing supplementary insight into the topic by analyzing transport modes and mobility patterns of inhabitants of the Greater Paris region. Through this micro-social analysis, we show that the more out-of-home activities people do, the lower their average speed in the course of a day is, whilst the average distance to a destination also decreases. This can be explained by the way they choose their destinations. For activities which are meaningful to them, they maximize the utility of their destination by using fast transport modes. For ancillary activities, they minimize travel time by locating them in the direct vicinity of their high-value activities, which prompts the use of slower modes. Finally, we show that the combination of these two rationales is easier in the center than in the suburbs, thus facilitating a high pace of life and more improvised schedules in the former. Hence, more than access to speed, it is residential location which induces disparities between people in their ability to deal with time demands

The subject of this thesis is entrepreneurship policy and the focus is on public support programmes directed to early stages technology-based ventures. It advocates that a broad view should be taken with regard to the type of policies for entrepreneurship that aim at supporting the facilitation of the supported ventures’ ability to link to their surrounding innovation system, in which they can secure their access to crucial external resources. Taking the departure in the perspective that early stages technology-based ventures are vulnerable, this thesis shows that publicly financed entrepreneurship supporting programmes can be motivated because small and newly started ventures have got limited resources, and because the risks, with which they are associated, generally make them unattractive in the eyes of private actors. Among the practical implications of this thesis, the following findings are emphasised: Public support, directed to the very earliest stages of venture development benefits from broadness, for the latter stages, specialisation and a higher degree of selectiveness could be an advantage. Moreover, it is important that the support provided take into account the whole process of business development, i.e. both the issues concerned with product/service and the issues connected to the market. A long-term commitment, cooperation between different actors and information about what support is available, are other aspects that are important for policymakers to consider in their design of public policies to support entrepreneurship. Finally, the importance of evaluative awareness must be emphasised from the very start. / Den här avhandlingen är en sammanläggning av 7 studier kring vilka en kappa med titeln ”Entreprenörskapspolicy: offentligt stöd för teknikbaserade företag” har skrivits. De inkluderade studierna beskrivs kortfattat i slutet av den här sammanfattningen. Syftet med avhandlingen är att analysera skälen till varför statliga interventioner kan vara berättigade, vidare att analysera målsättningar, design och resultat av offentliga program för stöd av nya teknikbaserade företag/projekt och slutligen att identifiera såväl praktiska som forskningsintressanta implikationer för utformningen av framtida stödaktiviteter. Avhandlingen baseras till största delen på olika typer av kvantitativa analyser av Stiftelsen Innovationscentrum (SIC) (studie 2-6). Förutom detta ingår en konceptuell studie där ett ramverk för att utvärdera inkubatorer, främst ur ett ”best practice” perspektiv, har tagits fram (studie 1). Slutligen har jag inkluderat en pågående uppföljningsstudie av Vinnovas program Vinn Nu (studie 7). Avhandlingen förespråkar en bred syn på begreppet entreprenörskapspolicy, där syftet med stödet är att underlätta för företag/projekt att koppla upp sig mot relevanta innovationssystem där de kan säkra sin tillgång på externa resurser. Med utgångspunkt i ett sårbarhetsperspektiv visar avhandlingen att offentligt finansierade program som stödjer entreprenörskap kan motiveras genom att de små och nystartade företagens resurser är begränsade och eftersom riskerna ofta gör att intresset från privata aktörer att gå in i dessa företag är lågt. Avhandlingens praktiska implikationer är bland annat att det stöd som riktas till de allra tidigaste utvecklingsfaserna i syfte att kvalificera idéer bör vara brett. Däremot kan ett mer selektivt urval vara fördelaktigt i något senare utvecklingsskeden och för mer specialiserade stöd. Vidare visar de inkluderade studierna på behovet av att se till hela företagets utvecklingsprocess. Det innebär bland annat att det är nödvändigt att kombinera hårt och mjukt stöd och att lika vikt läggs vid produkt- och marknadsutveckling. Offentligt stöd bör också vara långsiktigt, och samarbete mellan olika stödprogram är viktigt, inte minst med avseende på information om vilka stöd som finns och hur, var, när och av vem dessa kan sökas. Slutligen understryks vikten av att stödprogram redan från början bör ta hänsyn till utvärderingsaspekter.

In a hospital environment the focus is on quality patient care. One of the important factors to improve the quality of patient care is the implementation of a continuous professional development programme. In a continuous professional development programme of a hospital three stakeholders namely management, clinical facilitators and the nurse practitioners are identified as partners. When the stakeholders reach a point where it is believed that the continuous professional development programme does not serve its purpose to improve the quality of patient care it is time to evaluate and refine the programme. This study evaluate the critical care and high care units in a private hospital in Gauteng’s continuous professional development programme and the conclusions and recommendations are therefore limited to this specific clinical setting. The refinements recommended cannot be utilised in other clinical setting but the evaluation and research methodology can be utilised to evaluate similar continuous professional development programmes. Evaluation of a continuous professional development has the potential to negatively influence the nurse practitioners attitude towards continuous professional development. The objective of the research approach chosen namely a 4-D Appreciative Inquiry approach was to gain the collaboration of all the stakeholders in an effort to ensure that the evaluation of the continuous professional development programme of the critical care and high care units in the private hospital in Gauteng does not negatively affect the delivery of quality patient care. The Appreciative Inquiry research approach with its positive approach lends it to be the most appropriate research methodology of the study. The study utilised a Focus Group Inquiry as data a collection mechanism. From the study In-service training emerged as the element that form an important element of an effective continuous professional development programme but that it is of the at most importance that an effective Workplace learning environment exists. The supportive role of Management for a continuous professional development programme also emerged as one of the main themes required for the implementation of a continuous professional development programme. These findings were utilised in the refinement for the continuous professional development programme for the critical care and high care units in the private hospital in Gauteng. The research recommends that a steering committee be appointed to implement a pilot continuous professional development programme that incorporate the refinement elements identified. Due to the continuous improvement nature of the Appreciative Inquiry methodology it is also recommended that the pilot continuous professional development programme been re-evaluation to ensure long-term success and improvement. / Dissertation (MCur)--University of Pretoria, 2013. / gm2014 / Nursing Science / unrestricted

This study provides an empirical picture of the ways in which the New Public Management doctrine has been implemented in the context of post-Soviet states in Central Asia. Specifically, the data present evidence on the extent of school autonomy along six dimensions in the three states. The implications of the shift towards education decentralisation have been studied and explained. Based on a mixed method this evidence is drawn from three sources: surveys, interviews, and legislative analysis. With the research limitations in mind, based on the analysis undertaken, it is possible to identify some important messages regarding academic theory and education management practice. The first message is that neo-institutional theory, particularly historical institutionalism, has been supported by the evidence from the post-Soviet states. Despite quite diverse trajectories of the countries' political, economic and public sectors development over the last two decades, the policies promoted in education appear to be converging in both outputs and outcomes across the region. The three countries demonstrated persistent path-dependency through their inability to overcome institutional inertia, so that operational policy and structure dimensions have not been decentralized, combined with the effect of declining fiscal and bureaucratic capacity at the centre, so that managerial matters have been delegated to a school level. This trend was regional. The second message is that, the patterns observed in the three post-Soviet states displayed similarities to the patterns observed in education systems of the developed western democracies. Hence, the view that the structure of the post-communist welfare states is problematic, and that the particularities of their transition with budget cuts on top of the communist legacy and a hodge-podge of different approaches do not allow these states to be classified (cf. Orenstein 2008), has not been supported by the evidence. Education institutions in the three post-Soviet countries investigated conformed to a general West European pattern, although for different reasons.

In this research, carbon emissions and carbon savings in the smart grid are modelled and quantified. Carbon emissions are defined as the product of the activity (energy) and the corresponding carbon factor. The carbon savings are estimated as the difference between the conventional and improved energy usage multiplied by the corresponding carbon factor. An adaptive seasonal model based on the hyperbolic tangent function (HTF) is developed to define seasonal and daily trends of electricity demand and the resultant carbon emissions. A stochastic model describing profiles of energy usage and carbon emissions for groups of consumers is developed. The flexibility of the HTF for modelling cycles of energy consumption is demonstrated and discussed with several case studies. The analytical description to determine electricity grid carbon intensity in the UK is derived, using the available fuel mix data from the Elexon portal. The uncertain realisation of energy data is forecasted and assimilated using the ensemble Kalman filter (EnKF). The numerical optimisation of carbon emissions and savings in the smart grid is further performed using the ensemble-based Closed-loop Production Optimisation Scheme (EnOpt). The EnOpt involves the optimisation of fuel costs and carbon emissions (maximisation of carbon savings) in the smart grid subject to the operational control constraints. The software codes for the based on the application of EnKF and EnOpt are developed, and the optimisation of energy, cost and emissions is performed. The numerical simulation shows the ability of EnKF in forecasting and assimilating the energy data, and the robustness of the EnOpt in optimising costs and carbon savings. The proposed approach addresses the complexity and diversity of the power grid and may be implemented at the level of the transmission operator in collaboration with the operational wholesale electricity market and distribution network operators. The final stage of work includes the quantification of carbon emissions and savings in demand response (DR) programmes. DR programmes such as Short Term Operating Reserve (STOR), Triad, Fast Reserve, Frequency Control by Demand Management (FCDM) and smart meter roll-out are included, with various types of smart interventions. The DR programmes are modelled with appropriate configurations and assumptions in power plants used in the energy industry. This enables the comparison of emissions between the business-as-usual (BAU) and the smart solutions applied, thus deriving the carbon savings. Several case studies involving the modelling and analysing DR programmes are successfully performed. Thus, the thesis represents novel analytical and numerical techniques applied in the fast-growing UK market of smart energy solutions.

In recent decades, research from across the social sciences has demonstrated a strong, consistent and causal link between unemployment and a wide range of negative outcomes. These outcomes go beyond economic problems, incorporating issues such as low well-being, poor health and weak social capital. During the same time, successive UK governments have expanded the use of active labour market programmes (ALMPs): a wide range of interventions that aim to move unemployed people closer to the labour market. ALMPs have been widely evaluated since becoming a central part of UK social policy, yet the majority of studies focus almost exclusively on economic outcomes, such as re-employment and wage levels. This is despite the weight of evidence suggesting unemployment is as much a social problem as an economic one. This discrepancy has led to a small but growing body of research suggesting that ALMPs might play a role in modifying some of the health and social costs of unemployment: beyond simply moving people closer to the labour market. Using a mixed methods research design, this study examines whether ALMPs achieve this by considering four key questions. First, are ALMPs associated with higher well-being, health and social capital compared to the alternative of 'open unemployment'? Second, if there is an association, how robust is this and is there any evidence of a causal function? Third, does the context of an ALMP - such as the specific type of scheme and the kind of participant - matter for understanding outcomes? And fourthly, how and why do people's experiences of unemployment and ALMPs shape their health and well-being? The findings presented in this thesis offer five original contributions to the study of the health and social effects of ALMPs. First, there is a dichotomy in the effects of ALMPs: participants have higher well-being than the openly unemployed but similar health and social capital levels. Second, ALMPs are most effective in changing how participants feel about and evaluate their lives but are largely unsuccessful in mitigating negative emotions like anxiety. These two findings are evident in both cross-sectional and longitudinal data, suggesting the possibility of a causal function of ALMPs. Together, the findings suggest that the positive well-being effects of ALMPs are not necessarily linked to improved health or social capital but because participants begin to think about their lives in a different, more positive way. Third, well-being gains are experienced by both short-term and long-term unemployed people but disappear upon re-employment. This finding has an important implication for policy, with ALMPs seemingly effective as a short-term protective well-being measure. Fourth, this is the first UK study to explore whether ALMPs work more effectively for different types of unemployed people. The findings presented in Chapter Seven show that work-oriented ALMPs are more successful than employment-assistance programmes, whilst men, younger people, those with fewer qualifications, lower occupational status and lower pre-programme well-being experience the largest benefits of participation. Fifth, the qualitative analysis presented in Chapter Eight argues that ALMPs worked best when schemes reversed the perceived ‘losses’ associated with unemployment. Three processes of loss were identified - agency loss, functional loss and status loss – which, it is contended, help explain both the observed effects of ALMPs and the broader experience of unemployment. The thesis concludes with policy suggestions for improving the capacity of ALMPs to mediate the experience of unemployment.

L’objectif de ma thèse est d'étudier la propagation des ondes acoustiques dans les milieux poreux. La théorie de l'homogénéisation (Boutin et Auriault, 1990; Malinouskaya, 2007; Li, 2010) est utilisée avec les modèles de réseaux tels que LBM, LSM, LSM2S, LBM-LSM, LBM-LSM2S. Ces modèles nous permettent de déterminer les propriétés macroscopiques, les vitesses acoustiques et les effets d'atténuation dans les échantillons Fontainebleau avec deux composants (pores et quartz) et les échantillons STATOIL avec trois composantes (pores, quartz et d'argile). La modélisation numérique vise à résoudre 3 problèmes. Le premier problème est la caractérisation des échantillons; par la détermination de la porosité et des fonctions de corrélation avec les composantes de Fourier correspondantes (Adler, 1992; Nguyen, 2013). Le second porte sur la propagation des ondes dans les échantillons secs; les vitesses sont dérivées du tenseur de rigidité efficace C(eff) qui peut être calculé par LSM (Pazdniakou, 2012) ou LSM2S. Le troisième problème correspond aux échantillons saturés par un fluide incompressible ou compressible; les vitesses sont obtenues par résolution de l'équation de Christoffel après les déterminations de C(eff), fr la perméabilité dynamique K et ses réactions à la pression de fluide α et β. Pour les échantillons de Fontainebleau, les calculs sont effectués avec des modèles pré-existents tels que LSM, LBM, LBM-LSM. Ces modèles de bases sont étendus pour milieux avec plusieurs composants solides; ils sont validés via des comparaisons avec d’autres méthodes (Nemat-Nasser et Iwakuma, 1982; Torquato, 1998, 2000; Cohen, 2004). Les vitesses, le module d’élasticité et cisaillement efficace de tous les échantillons secs ainsi que les vitesses et l'atténuation des ondes dans les échantillons saturés sont déterminés. Les séries de résultats obtenues concordent bien avec des corrélations empiriques et théoriques, tels que le modèle d’IOS d’ Arns (1996), les modèles empiriques de Nur et al. (1991), Krief (1990) et avec le modèle de Gassmann. Les résultats numériques sont un peu plus grands que les données expérimentales d’ Han (1986) et de Gomez et al. (2010); les raisons de cette petite différence ont été provisoirement analysées, mais sa cause n'a pas été identifiée sans ambiguïté. / The purpose of this Ph.D. thesis is to study acoustic waves in porous media. The homogenization theory (Boutin and Auriault, 1990; Malinouskaya, 2007; Li, 2010) is used together with the lattice models such as LBM, LSM, LSM2S, LBM-LSM, LBM-LSM2S in order to determine the macroscopic properties, the acoustic velocities, the attenuation effects in Fontainebleau samples with two components (pore and quartz) and in STATOIL samples with three components (pore, quartz and clay). Three problems are studied numerically in this work. The first problem is devoted to characterizations of samples; this is done with the determination of the porosity and of the correlation functions with the corresponding Fourier components (Adler, 1992; Nguyen, 2013). The second one addresses wave propagation in dry samples; the velocities are derived from the effective stiffness tensor C(eff) which can be calculated by LSM (Pazdniakou, 2012) or LSM2S. The third one corresponds to samples saturated by incompressible or compressible fluids; the velocities can be obtained from the Christoffel equation after determining C(eff) , the dynamic permeability K and the reactions to fluid pressure α and β. For Fontainebleau samples, the calculations are performed with basic existing models such as LSM, LBM, LBM-LSM. These basic models are extended to solids with multiple components; they are validated by comparisons with others (Nemat-Nasser and Iwakuma, 1982; Torquato, 1998, 2000; Cohen, 2004). The velocities, the effective bulk and shear modulus of all the dry samples as well as the velocities and the attenuation effected in saturated samples are determined. These results are in good agreement compared with existing models and results, such as the IOS model of Arns (1996), the empirical models of Nur et al. (1991), Krief (1990) and with Gassmann’ s model. The numerical results are slightly larger than the experimental data of Han (1986) and Gomez et al. (2010); the origin of this small discrepancy has been tentatively analyzed, but its cause has not been unambiguously identified.

Magister Educationis - MEd / The main aim of the study was to investigate change of programmes of study at University of the Western Cape (UWC). This involved examining the extent, nature and possible range of factors which shape change of programmes. The factors influencing choice to study at UWC and choice of programmes formed the background for understanding the link between choice and change of programmes. The study focused on first-time entering undergraduates in selected faculties: Arts, Economics and Management Sciences (EMS) and Community and Health Sciences (CHS). The progression “pathways†of these students were tracked from first year of enrolment in 2001 to 2004. The findings of the longitudinal data showed that the percentage of students who changed their programmes in the Faculty of Arts and in CHS was very minimal. In total 5% changed their programmes in CHS in three-year programmes while in four-year programmes only 2% changed their programmes. In the Faculty of Arts in three-year and four-year programmes change of programmes ranged from 6% to 10%. The significant change of programmes of study happened in EMS which also had the highest enrolments. In total 18% of the students changed their programmes. It appeared from the statistical data that failure and academic exclusion could be associated with change of programmes of study. Other factors such as financial problems, poor grouping of courses, exploring and changing to preferred programmes as well came out from the interviews as major influences of change of programmes. / South Africa

Magister Educationis - MEd / Although there are several achievements made by the Namibian government after independence, there are still many constraints and challenges facing the education system, particularly with regard to novice teachers' support and professional development. Most novice teachers are coming into the teaching profession and receive little or no assistance. Nevertheless, novice teachers are expected to perform the full teaching responsibilities despite their inexperience. Novice teachers are expected to formulate their own coping strategies and to grapple with the challenges they encounter during their first year of teaching on their own. This study aims at investigating the coping mechanisms used by novice teachers to achieve their teaching goals during their first year of teaching. The study recommends that the teacher training institutions, the Ministry of Education and schools have an important role to play in facilitating the use of effective coping mechanisms and in assisting novice teachers in alleviating these challenges. Teacher training should ensure that a balance is maintained between theory and practice. In addition the Ministry needs to provide favourable working conditions and put up support systems for novice teachers. Schools should also provide orientation to all novice teachers and most especially render support to all novice teachers until they are fully integrated into the school and teaching culture. / South Africa