La préférence homosexuelle masculine (PHM) est considérée comme un paradoxe darwinien car c'est un trait partiellement héritable, induisant une baisse de succès reproducteur, ayant une fréquence (en Occident) entre 2 et 6 % et qui semble propre à l'espèce humaine (en excluant les espèces domestiquées). La sélection de parentèle, ou bien un facteur pléiotrope antagoniste sont classiquement proposés, bien que l'essentiel des données proviennent de sociétés occidentales. Nous avons proposé une nouvelle hypothèse, basée sur la stratification sociale et permettant d'expliquer la spécificité humaine, dans laquelle la migration sexe-dépendante de type hypergyne amplifie la sélection d'un gène antagoniste. La plausibilité de cette hypothèse est confirmée par des modèles théoriques d'ESS et des modélisations individu-centrés. De plus, une analyse comparative basée sur données anthropologiques a permis de montrer un lien entre le niveau de stratification d'une société et la probabilité que la PHM soit connue dans cette société. La sélection de parentèle, mesurée dans une société non-occidentale (Indonésie), semble participer à la compensation du coût reproductif de la PHM, mais insuffisamment pour en expliquer le maintien. Les principales caractéristiques familiales associées à la PHM, l'excès de grands frères, et la fécondité accrue dans certaines classe d'apparentés, se retrouvent également en Indonésie. Toutefois, en Indonésie, les classes d'apparentés concernées différent en partie des résultats obtenus en occident et ne correspondent pas aux attendues de l'hypothèse de sélection sexuellement antagoniste portée par le chromosome X. La prise en compte de l'interaction entre facteurs sociaux et biologiques ainsi que l'acquisition de données hors des sociétés occidentales ouvrent de nouvelles perspectives de compréhension de ce paradoxe darwinien. / Male homosexual preference (MHP) is considered as a Darwinian paradox, as it is partially heritable, induces a reproductive cost, has a prevalence estimated (in western countries) between 2 and 6%, and seems to be specific to humans (excluding domesticated species). Kin selection or sexually antagonistic factors have been proposed as possible explanations, although most data come from western countries. Here, we proposed a new hypothesis, based on social stratification, able to explain the specificity of human MHP, in which hypergyny (up-migration of the women) enhances the selection of a sexually antagonistic gene. This hypothesis is supported by theoretical models (ESS) and individual-based modelling. Furthermore, a comparative analysis based on anthropological data showed that the level of social stratification predicts the probability for MHP to be known in a society. Kin selection was tested in a non-western society (Indonesia), and seems to contribute to the compensation of the reproductive costs associated with MHP, although this effect alone was insufficient to explain the maintenance of MHP. The main biodemographic features associated with MHP, an excess of older brother and an increased fecundity of some classes of relatives, were also found in Indonesia. However, the classes of relatives affected by this increased fecundity differ from the classes affected in western populations, and differ from the predictions of the sexually antagonistic genetic factor hypothesis displayed on an X chromosome. Taking into account the interaction between social and biological factors, together with the acquisition of new data in non-western populations, open new perspectives in our understanding of this Darwinian paradox.

La combustion stratifiée, qui consiste à brûler un mélange carburant/oxydant inhomogène, et la combustion diluée, consistant à ajouter une quantité limitée de gaz brûlés, sont deux technologies utilisées dans les moteurs à piston pour réduire leur consommation. Cette thèse est dédiée à l’étude de l’allumage dans ces deux types de milieux en régimes laminaire et turbulent. Un nouveau schéma cinétique pour la combustion propane/air a été dérivé et combiné à deux approches de modélisation différentes : la chimie complexe et une approche de chimie tabulée de type FPI. Dans le cas laminaire, les deux approches de modélisation donnent des résultats similaires et un modèle simple a mis en évidence l’importance de la dynamique des gaz frais et des gaz brûlés sur le développement du noyau. Dans le cas turbulent, plusieurs techniques d’analyse ont montré la dépendance de la vitesse absolue de la flamme au champ de vitesse moyen et la décorrélation des fluctuations locales de richesse. / Stratified combustion, which consists in burning an inhomogeneous fuel/air mixture, and diluted combustion, which consists in adding a limited quantity of burnt gases, are two technologies used in internal combustion engines to reduce fuel consumption. This Ph.D is devoted to the study of ignition in these two types of combustion in laminar and turbulent regimes. A new kinetic scheme for propane/air combustion has been derived and combined to two modeling approaches: finite-rate chemistry and an FPI tabulated chemistry approach. In the laminar case, both approaches give similar results and a simplified model has highlighted the importance of fresh and burnt gases dynamics on the kernel development. In the turbulent case, several techniques of analysis have shown the dependency of absolute flame speed on the mean fluid velocity and the lack of correlation to the local equivalence ratio.

Cette recherche doctorale examine, dans un pays en proie à une « passion pour l’égalité » (Forsé et al, 2013), le degré d’acceptation à l’égard des très hauts revenus. Elle se situe au croisement de la sociologie des inégalités, de l’économie du vedettariat et de la justice sociale.Déjà présente dans le débat opposant John Rawls (1971 et 2001) et Robert Nozick (1974) au travers de l’exemple de Wilt Chamberlain, la figure de la star nous permet de répondre à une question centrale et contemporaine : les très hautes rémunérations sont-elles justes socialement ? La thèse se base sur un travail qualitatif articulé autour de la réalisation et de l’analyse de 55 entretiens semi-directifs. Elle repose également sur un travail d’analyse quantitative qui a permis de représenter l’évolution des inégalités de revenus ainsi que l’évolution des revenus des sportif.ve.s et des comédien.ne.s les mieux payé.e.s en France et dans le monde. Des régressions linéaires ayant pour variables à expliquer les rémunérations des stars de football ou de cinéma permettent de mesurer l’effet propre des caractéristiques individuelles des stars.La première partie détermine la place occupée par les comédien.ne.s et les footballeurs professionnels, désigné.e.s par la presse comme stars, dans la distribution des revenus en France. La deuxième partie élabore une définition subjective du statut de star et restitue les déterminants, selon les personnes interrogées, de la rémunération des stars. L’analyse des réponses des personnes interrogées à propos des facteurs explicatifs des revenus des cadres d’entreprise permet de mettre en avant la spécificité des stars. La troisième partie analyse l’opinion des Français à l’égard des revenus des stars. Inspirée de l’enquête PISJ (Forsé et Galland, 2011), une liste de 10 métiers – ou statuts — appartenant au décile (voire pour certains au centile ou au millime) supérieur (star de cinéma, star de football, blogueur, mannequin, animateur de télévision, médecin généraliste, professeur d’université, ministre, cadre supérieur et PDG d’une grande entreprise) a permis de constituer un dispositif empirique destiné à apprécier le degré d’acceptation à l’égard des rémunérations des stars dans une réflexion élargie aux plus hautes rémunérations. Il en résulte que le niveau de la rémunération de Cristiano Ronaldo, de Lionel Messi, d’Angelina Jolie et de Gérard Depardieu n’est accepté ni en raison d’une justification libertarienne (Nozick, 1974) ni par rejet de l’arbitraire du mérite individuel (Rawls, 1971 et 2001). Les très hauts revenus — et notamment ceux des stars — font l’objet d’une acceptation qui s’explique majoritairement par la combinaison d’un principe de stratification (logique individuelle de contribution-rétribution) et d’un principe de correction (l’utilité sociale). Enfin, l’examen des variables sociodémographiques montre que plus la personne se situe à gauche du spectre politique, plus elle critique les très hautes rémunérations. L’acceptation des très hauts revenus est donc un jeu à trois pôles entre une logique économique (contribution individuelle-rétribution individuelle), un principe de correction (l’utilité sociale) et un attachement à la notion d’égalité (le positionnement politique). / Drawing on the sociology of inequality, the star system economy and social justice, this PhD dissertation aims to examine whether French people, who live in an “equality-obsessed country“ (Forsé et al. 2013), accept high earnings.The “star”, which has already been used in the debate between Rawls (1971; 2001) and Nozick (1974) through the example of the well-known NBA star-player Wilt Chamberlain, enables us to answer one key and contemporary question: do people believe that high earnings are socially fair? To investigate this issue, I used both qualitative and quantitative methods, working with a total sample of 59 interviewees and conducting 55 semi-structured interviews, each two-hours long on average. The quantitative part of my research consists of a statistical analysis using multiple data sources (INSEE, WID, annual pundit surveys…), in order to provide an account of the evolution of movie and football stars’ earnings, in relation to the general evolution of incomes and incomes in the top percentiles (top 1%, top 0.1%, top 0.01%, and top 0.001%). Regression linear analyses allow us to determine the impact of the stars’ individual characteristics.The aim of the first part of this dissertation is to find how the top paid football players and actors (dubbed as football stars and movie stars by the press) are ranked in France’s income distribution. The second part seeks to provide a subjective definition of the star’s status and to identify, according to the interviewees, how their earnings are garnered. The analysis of the factors explaining the earnings of management executives by the interviewees helps us to insist on the specificity of the stars’ earnings. The third part examines French opinion on stars’ earnings. Using a PISJ-inspired list (Forsé et Galland, 2011) of 10 jobs or statuses belonging to the 10% (or top 1% or top 0.1%) (movie star, football-star, blogger, model, TV host, doctor, university teacher, management executive), I tried to investigate empirically the interviewees’ attitude towards not only stars’ earnings but high earnings in general.Finally, my research shows that interviewees accept the position of Cristiano, Lionel, Angelina Jolie and Gérard Depardieu among top earners, and agree with their earnings exceeding millions of euros. The joint use of the two principles (on the one hand, the stratification principle, based on the acceptance of an economics-based logic, and on the other hand, the corrective principle of social utility) leads to a reasoned acceptance of the earnings of movie stars and football stars. I therefore show that the acceptance of the level of movie stars’ and football stars’ earnings differs from the libertarian approach of Robert Nozick (1974) and from the rejection of individual merit (Rawls, 1971 and 2001). The analysis of the interviewees’ answers in relation to sociodemographic variables indicates that there is a link between the interviewees’ political beliefs and their attitude toward high earnings. The more left-wing they declared themselves, the more critical they were of the perceived hierarchy in earnings and of the very high ones. The acceptance of high earnings can be seen as a ménage-à-trois between an economics-based logic (individual contribution leads to individual earning), a corrective principle (social utility) and the valuation of equality (political belief).

This study describes my experience 111 the research 1 conducted in two Harare secondary schools in which the interpretive ethnographic qualitative research methodology was used. The objective of the research was to assess the infuence of ability grouping on learners, given that ability grouping is a common practice in Zimbabwean schools. In particular, this study sought to find out how this practice affected classroom instruction, learner performance and the social relationship among learners. This study was conducted over three months at each of the schools, during which time formal interviews with teachers, administrators and learners, were conducted. These were complemented with informal conversations, where relevant comments were noted. Observation and limited participation in the two schools were also employed as means of collecting data. Analysis of documents was also done to supply more data about the practice and how it was implemented. As customary with qualitative research, I, as the researcher, was the instrument for data collection. Data were analysed by identifying themes which emerged from the statements of the respondents, and interpretation was done using the mechanism of thick description of what the respondents said and did during the interviews and observation, and what the documents had to say about this practice. I, as researcher, concluded from this study that ability grouping had a negative effect on classroom instruction for the low ability groups, since teachers tended not to prepare thoroughly for them. There was also a tendency among high ability groups to look down upon the low ability groups and to stigmatise them as "those who do not want to learn." Naturally this created a counter reaction from the low ability groups, making social relationships between members of the groups sour. l, as researcher, recommend a re-examination of this practice by the schools, with a view to either discontinue it or to work out mechanisms to remove the negative factors. / Educational Studies / D. Ed. (Didactics)

Recently, the issue of poverty amongst indigenous people has become a significant topic in literature on social policy and development studies. The literature mainly looks at this issue in terms of an unequal and one-sided relationship between the mainstream society and an indigenous minority group. This thesis seeks insights into the more diversified circumstances and experiences of poverty amongst indigenous people. The following questions are addressed: (1) Why and how is the poverty of indigenous people reproduced over time and space? (2) How can we understand patterns of differentiation between indigenous people? (3) What is the balance between structural opportunity and constraint in the lives of indigenous people? (4) To what extent do people exercise agency to cope with or overcome their poverty situations? The thesis adopts an ethnographic approach, including participant observation and interviews in four villages of Taiwanese indigenous people. It explores the connections between poverty dynamics and diversified patterns of socio-economic action amongst indigenous people. Hogget and Greener's model of agency, which contains the essential theoretical views of Giddens (the ability of agents to act) and Bourdieu (the embedded corporeal disposition of human agents) are used to make sense of this exploration. The thesis finds that the actions of indigenous people as human agents are differentiated. Actions are not only motivated by strategic plans and emotions but are also influenced by the agents' socio-economic positions, such as their occupations and education and those of their parents. The differentiated socio-economic activities of agents, in turn, have a strong effect on the stratification of their living standards.

Turbulent flows affected by buoyancy lie at the basis of many applications, both within engineering and the atmospheric sciences. A prominent example of such an application is the atmospheric boundary layer, the lowest layer of the atmosphere, in which many physical processes are heavily influenced by both stably stratified and convective turbulent transport. Modelling these turbulent flows correctly, especially in the presence of stable stratification, has proven to be a great challenge and forms an important problem in the context of climate models. In this thesis, we address this issue considering an advanced class of turbulence models, the so-called explicit algebraic models.In the presence of buoyancy forces, a mutual coupling between the Reynolds stresses and the turbulent heat flux exists, which makes it difficult to derive a fully explicit turbulence model. A method to overcome this problem is presented based on earlier studies for cases without buoyancy. Fully explicit and robust models are derived for turbulence in two-dimensional mean flows with buoyancy and shown to give good predictions compared with various data from direct numerical simulations (DNS), most notably in the case of stably stratified turbulent channel flow. Special attention is given to the problem of determining the production-to-dissipation ratio of turbulent kinetic energy, for which the exact equation cannot be solved analytically. A robust approximative method is presented to calculate this quantity, which is important for obtaining a consistent formulation of the model.The turbulence model derived in this way is applied to the atmospheric boundary layer in the form of two idealized test cases. First, we consider a purely stably stratified boundary layer in the context of the well-known GABLS1 study. The model is shown to give good predictions in this case compared to data from large-eddy simulation (LES). The second test case represents a full diurnal cycle containing both stable stratification and convective motions. In this case, the current model yields interesting dynamical features that cannot be captured by simpler models. These results are meant as a first step towards a more thorough investigation of the pros and cons of explicit algebraic models in the context of the atmospheric boundary layer, for which additional LES data are required. / <p>QC 20150522</p>

During the 1960s and early 1970s, the black – white wage gap narrowed significantly, but has remained constant since the late 1980s. The black – white wage gap in the recent period may reflect differences in human capital. A key component of human capital is labor market experience. The first chapter of this dissertation examines how differences in the returns and patterns of experience accumulation affect the black – white wage gap. Accounting for differences in the nature of experience accumulation does not explain the very large gap in wages between blacks and whites. Instead, the wage gap seems to be driven by constant differences between blacks and whites which may represent unobserved differences in skill or the effects of discrimination. The second chapter of the dissertation examines the role of discrimination in explaining the wage gap by asking whether statistical discrimination by employers causes the wages of never incarcerated blacks to suffer when the incarceration rate of blacks in an area increases. I find little evidence that black incarceration rates negatively affect the wages of never incarcerated blacks. Instead, macroeconomic effects in areas with higher incarceration rates play a more important role in explaining the variation in black wages. The third and final chapter of the dissertation examines the black – white wage gap and its determinants across the entire wage distribution to determine if the factors that are driving the wage gap vary across the distribution. I find that at the top of the conditional distribution, differences in the distribution of characteristics explain relatively more of the black – white wage gap than differences in the prices of characteristics. At the bottom of the conditional distribution, differences in the distribution of characteristics explain relatively more of the wage gap—although this finding varies across different specifications of the model.

旅遊風氣在台灣日盛且趨於普遍，且是重要的消費活動現象。台灣的旅遊活動支出在2001年已經高達5780億元，休閒渡假風氣日漸普及。而過去的旅遊觀光研究多以行銷、或是觀光景點的規劃為主要目的，以社會學角度進行研究者在國外雖已有所累積，在中文世界裡仍屬少數。因此，本論文以社會學中階層品味區分的角度探討台灣的旅遊活動，針對2001年交通部觀光局所做的「國人旅遊狀況調查」進行量化分析的二手資料研究，並輔以歷史資料的縱貫面分析。透過上述材料，討論不同階層間的旅遊活動品味差異，以及旅遊發展過程中的階層化過程。 在研究中，以階層化作為主要的分析主軸。在量化資料中，依因素分析區分出四類的旅遊喜好：都市旅遊、文化旅遊、大眾自然旅遊與深度自然旅遊。這四類型的旅遊分別進行迴歸分析，統計發現，經濟資本的收入變項並未發揮品味的影響力，世代性的年齡因素以及正式教育制度的文化資本則能解釋旅遊喜好的品味差異。回顧台灣的旅遊活動脈絡發展，透過制度上的發展以及群體間的運作，產生了因文化資本差異所形成的階層差異，旅遊活動的意義不只是個人主觀心理層次的好惡，亦可反映出客觀階層的群體效果。是故，本研究解釋不同階層者選擇不同旅遊活動的原因，另檢證學者所提出的品味理論以及旅遊社會學的相關分析，提出台灣旅遊活動的階層化發展過程以及目前的品味區分現象，說明台灣旅遊現象的發展以及現況。 / More and more people in Taiwan spend lots of time going traveling and tourism gradually becomes one of the most important consumption activities. In 2001, total fee of citizens’ traveling expenses was 5780 hundred millions Taiwan dollars. The fact shows that tourism is an influential social phenomenon and we should understand it as deliberately as possible. Actually, lots of researches about tourism in the perspective of sociology have been published in the world but there is few in Taiwan. Therefore, this thesis takes a sociological approach that adopts the theory of hierarchy and distinction of taste. This research used the quantitative analysis with the sample data collected by Tourism Bureau in 2001 and the qualitative analysis with historical data. This study describes the process of stratified and stratifying differentiation of tourism in Taiwan. Four types of tourism , urban tourism、cultural tourism、scenic tourism and deep scenic tourism are found in the data through factor analysis. The data demonstrated that cultural capital, not economical capital effects people choosing types of tourism in the regression model. The difference of cultural capital in this thesis represents the difference between age、educational capital and occupation of people. Consequently, the choice of the type of tourism is not only a casual or random choice, but a choice with the stratification of the society. Finally, this research examined the theory about taste and sociology of tourism to combine with the results of quantitative and qualitative analysis.

Multicollinearity can be present in the propensity score model when estimating average treatment effects (ATEs). In this thesis, logistic ridge regression (LRR) and principal components logistic regression (PCLR) are evaluated as an alternative to ML estimation of the propensity score model. ATE estimators based on weighting (IPW), matching and stratification are assessed in a Monte Carlo simulation study to evaluate LRR and PCLR. Further, an empirical example of using LRR and PCLR on real data under multicollinearity is provided. Results from the simulation study reveal that under multicollinearity and in small samples, the use of LRR reduces bias in the matching estimator, compared to ML. In large samples PCLR yields lowest bias, and typically was found to have the lowest MSE in all estimators. PCLR matched ML in bias under IPW estimation and in some cases had lower bias. The stratification estimator was heavily biased compared to matching and IPW but both bias and MSE improved as PCLR was applied, and for some cases under LRR. The specification with PCLR in the empirical example was usually most sensitive as a strongly correlated covariate was included in the propensity score model.

Educational transitions in the UK are related to social background characteristics such as social class and, to a lesser extent, ethnicity and gender. This thesis presents a case study of admission to the University of Oxford to understand why, conditional on application, admissions patterns into selective higher eduction in Britain show an advantage for already privileged strata of society. Specifically, net of attainment, the professional middle class, white, male and state school applicants fare particularly well in securing offers for undergraduate study at Oxford. With the exception of the state school effect, the admissions privilege advantages already privileged strata of society. In the first empirical section, the analysis of purposefully generated survey data on 1,929 applicants for admission to the University of Oxford finds that quantifiable measures of merit fail to fully explain differential admissions patterns. The logistic regression models also uncover that while applicants from the private sector initially have similar gross chances of gaining an offer to their state school educated peers, they actually face a penalty in the selection process when taking into account their higher levels of prior academic attainment. Furthermore, the analysis shows that while measures of cultural capital, motivation, aspiration and learning style are meaningfully related to selection decisions, they do not explain the lower transition rates for ethnic minority applicants, those from non-professional class backgrounds, female applicants and private school applicants. The second step in the empirical investigations then aims to understand the generative mechanisms behind these findings from the perspective of the decision makers in the selection process. This section draws on interviews with 25 admissions tutors and the observation of eight admissions meetings. The analysis here finds that selectors view the admissions exercise as involving risks and uncertainties. Also, many participating tutors routinely considered schooling in their selection decisions and discounted the performance of applicants who had come from very high achieving schools but who were not top achievers within this peer group. The mechanism of homo-social reproduction in decisions involving uncertainty is then put forward as a possible explanation for the unequal transition patterns. Finally, the third empirical analysis section investigates links between degree performance in final university examinations and admissions relevant factors. This section includes the degree performance of Oxford students as well as those who subsequently embarked on their degree course at universities other than Oxford. The most striking finding is that among the Oxford graduates, female and private school students are less likely to achieve first class degrees than their male and state school educated peers. One interpretation of this finding is that the discounting that selectors apply in the admissions process for these applicants is not only justified but may not even go far enough. But it is also possible, in particular with regard to the female effect, that the Oxford study environment or the examination system, or both are more conducive to male achievements. This thesis contributes to sociological theory by showing that existing models of educational transition have paid insufficient attention to the role of gatekeepers and their individual preferences in generating aggregate selection patterns. Incorporating selectors as actors in transition models increases our understanding of unequal access to educational institutions and the challenges faced in striving towards equal opportunities in an education based meritocracy. The findings presented here have implications for other fields of sociological inquiry that need to account for the role of individual decision makers such as labour market research. The work presented here has implications for policy making regarding selection processes within the University of Oxford and British higher education more generally. It could also aid university systems such as Germany, that are moving towards selective admission, to think about the challenges of designing truly equitable selection processes.