Educational attainment has been a central debate in the field of educational research for a long time. Moreover, regarding inequality in educational attainment, social research has been dominated by questioning the association between educational attainment and gender, ethnic and social class inequalities. When considering social class inequality, one universal conclusion can be drawn: the higher family social status, the better academic attainment. This is also true in Taiwan, as is the case elsewhere. However, I want to ask: does social status influence other educational outcomes in addition to academic attainment, and if so, how? With respect to this question, I argue that beside educational attainment, students’ school experiences can be considered as an important educational outcome; furthermore, it may be influenced by social status and gender. The aim of this research is to investigate the experiences of year 8 students in Taiwan: what are their perceptions of school experiences?, and more specifically, what extent family social status and gender are associated with different students’ perception of parents’ educative capital and their own educational habitus, which in turn influences pupils’ school experience. Indeed, this study intends to expand the relationship between social status/gender and education; moreover, to examine a dynamic structure between family social status/gender influence and personal perception. After the process of data analysis, many meaningful findings are examined. Family socioeconomic status did not make direct impact on students’ relationship with peers and teachers; and parents’ educative capital did not affect educational attainment directly either. In addition, surprisingly, gender difference made no difference on all measured aspects. By holding such information, it will be possible examine the phenomenon of youth development and secondary education in Taiwan.

The high prevalence rates of child overweight and obesity within the UK is a serious problem, and one that has received a lot of attention from policy makers, researchers and the media. There is some evidence for socioeconomic inequalities in child overweight and obesity, with children in less advantaged socioeconomic groups at an increased risk of being overweight or obese. The nature of these inequalities is not well understood. Within this thesis I consider different aspects of socioeconomic status and their relationship with child overweight and obesity. There are three distinct strands of the investigation. Firstly, I consider whether socioeconomic inequalities in child overweight and obesity have changed over time. This is followed by two separate analyses of the relationship between obesity and overweight with parental income and education. I find evidence that socioeconomic inequalities in child overweight and obesity have widened over time, but only because of the relatively low increases in child overweight and obesity amongst children from the most advantaged families. I investigate whether there is an association between income and child overweight, and find that other parental characteristics, namely parental education, can explain correlations between familial income and child overweight and obesity. I find that father’s education has a stronger association with child overweight and obesity than does mother’s education, and that this is not solely because father’s education is a better predictor of the family’s financial and economic resources.

Class analysis suggests a strong positive link between parental social position and adolescent educational expectations. Yet, Indian, Pakistani, Bangladeshi and Black Caribbean pupils with disadvantaged parents maintain much higher expectations for university study compared to their white peers from more advantaged families. Ethnicity is key to understanding this paradox. Yet, quantitative research has not explained how ethnicity impacts on adolescent expectations or whose expectations it impacts most. This thesis investigates whether the longitudinal association between parental social position at age 14 and pupils expectations at age 16 is mediated by parent-child conflict, pupils’ homework engagement, feelings about school and pupils’ assessments of teacher effectiveness at pupils’ age 15, net of all prior influences at age 14. It then explores whether these mediational routes are moderated by maternal ethnicity. Ecological systems theory informs a longitudinal latent variable mediation model estimated on panel data from waves 1-3 of the Longitudinal Study of Young People in England. Parental social position at age 14 does not affect pupils’ expectations at age 16 via the hypothesised home and school factors at age 15. Its longitudinal influence is weak on white, modest on Black Caribbean and Pakistani and insignificant on Indian and Bangladeshi pupils’ expectations at age 16 casting doubt on classical sociological models. Moderation by maternal ethnicity strengthens the positive longitudinal influence of home-related factors on expectations and tones down the negative effect of family disadvantage or other negative influences particularly on South Asian pupils’ expectations at age 16. Pupils’ expectations at age 14 significantly impact on home and school factors and expectations at age 15 and are the primary indirect route of influence on expectations at age 16. Moderated by maternal ethnicity, this mediational mechanism allows South Asian pupils to develop and maintain higher expectations than white pupils despite family disadvantage.

Cette thèse contribue à l'étude théorique et a l'application de la vérification et de la synthèse quantitative. Nous étudions les stratégies qui optimisent la fraction de deux récompenses des MDPs. L'objectif est la synthèse de régulateurs efficaces dans des environnements probabilistes. Premièrement nous montrons que les stratégies déterministes et sans mémoire sont suffisants. Sur la base de ces résultats, nous proposons trois algorithmes pour traiter des modèles explicitement encodées. Notre évaluation de ces algorithmes montre que l'un de ces derniers est plus rapide que les autres. Par la suite nous proposons et mettons en place une variante symbolique basé sur les diagrammes de décision binaire.Deuxièmement, nous étudions le problème de réparation des programmes d'un point de vue quantitatif. Cela conduit à une reformulation de la réparation d'un log: que seules les exécutions fautives du programme soient modifiées. Nous étudions les limites de cette approche et montrons comment nous pouvons assouplir cette nouvelle exigence. Nous concevons et mettons en œuvre un algorithme pour trouver automatiquement des réparations, et montrons qu'il améliore les modifications apportées aux programmes. Troisièmement, nous étudions une nouvelle approche au framework pour la vérification et synthèse quantitative. La vérification et la synthèse fonctionnent en tandem pour analyser la qualité d'un contrôleur en ce qui concerne, par exemple , de robustesse contre des erreurs de modélisation. Nous considérons également la possibilité d'approximer la courbure de Pareto, qui appataît de la combinaison du modèle avec de multiples récompenses. Cela nous permet à la fois d'étudier les compromis inhérents au système et de choisir une configuration adéquate. Nous appliquons notre framework aux plusieurs études de cas. La majorité de l'étude de cas est concernée par un système anti-collision embarqué (ACAS X). Nous utilisons notre framework pour aider à analyser l'espace de conception du système et de valider le contrôleur en cours d'investigation par la FAA. En particulier, nous contribuons l'analyse par PCTL et stochastic model checking. / This thesis contributes to the theoretical study and application of quantitative verification and synthesis. We first study strategies that optimize the ratio of two rewards in MDPs. The goal is the synthesis of efficient controllers in probabilistic environments. We prove that deterministic and memoryless strategies are sufficient. Based on these results we suggest 3 algorithms to treat explicitly encoded models. Our evaluation of these algorithms shows that one of these is clearly faster than the others. To extend its scope, we propose and implement a symbolic variant based on binary decision diagrams, and show that it cope with millions of states. Second, we study the problem of program repair from a quantitative perspective. This leads to a reformulation of program repair with the requirement that only faulty runs of the program be changed. We study the limitations of this approach and show how we can relax the new requirement. We devise and implement an algorithm to automatically find repairs, and show that it improves the changes made to programs.Third, we study a novel approach to a quantitative verification and synthesis framework. In this, verification and synthesis work in tandem to analyze the quality of a controller with respect to, e.g., robustness against modeling errors. We also include the possibility to approximate the Pareto curve that emerges from combining the model with multiple rewards. This allows us to both study the trade-offs inherent in the system and choose a configuration to our liking. We apply our framework to several case studies. The major case study is concerned with the currently proposed next generation airborne collision avoidance system (ACAS X). We use our framework to help analyze the design space of the system and to validate the controller as currently under investigation by the FAA. In particular, we contribute analysis via PCTL and stochastic model checking to add to the confidence in the controller.

<p> Self-regulation is considered one of the most important operations of the human self and is correlated with multiple variables. The objective of this quantitative study was to explore the relationship between one&rsquo;s capacity to generate vivid visual mental images and one&rsquo;s capacity for both behavioral and physiological self-regulation. Fifty-three participants completed three assessments for this study: the Vividness of Visual Imagery Questionnaire (VVIQ-2) measured capacity for mental imagery; the Self-Regulation Questionnaire measured behavioral self-regulation; and a baseline assessment of heart rate variability using a HeartMath^&reg; device measured physiological self-regulation. When using bivariate correlations to analyze the data, the results revealed that a significant relationship existed between capacity for visual imagery and self-reported behavioral self-regulation. Using a more modest &alpha; level of 90%, there was also a weak relationship between capacity for visual imagery and the predetermined markers for physiological self-regulation. These results underscore the value of visual imagery and imagery vividness when working with individuals struggling with self-regulation and they open the door to exploring causality between the two. </p><p>

abstract: Daily dairies and other intensive measurement methods are increasingly used to study the relationships between two time varying variables X and Y. These data are commonly analyzed using longitudinal multilevel or bivariate growth curve models that allow for random effects of intercept (and sometimes also slope) but which do not address the effects of weekly cycles in the data. Three Monte Carlo studies investigated the impact of omitting the weekly cycles in daily dairy data under the multilevel model framework. In cases where cycles existed in both the time-varying predictor series (X) and the time-varying outcome series (Y) but were ignored, the effects of the within- and between-person components of X on Y tended to be biased, as were their corresponding standard errors. The direction and magnitude of the bias depended on the phase difference between the cycles in the two series. In cases where cycles existed in only one series but were ignored, the standard errors of the regression coefficients for the within- and between-person components of X tended to be biased, and the direction and magnitude of bias depended on which series contained cyclical components. / Dissertation/Thesis / M.A. Psychology 2013

<p> This quantitative study examines how attachment and masculinity influence men in their participation in social groups and support or therapy groups as measured by the ECR-RS (Fraley, Brumbaugh, Heffernan, &amp; Vicary, 2011) and the MRNI-SF (Levant, Hall, &amp; Rankin, 2013). An online survey was given to 308 U.S. male adults asking questions about their attachment to their primary partner, their family of origin, social groups they participate in, support or therapy groups they participate in, and their endorsement of traditional masculine gender norms. The results from this study suggest the following: that men will have the same level of attachment to their family of origin as they do to both romantic dyadic relationships and to social groups they participate in, men are more securely attached the more they participate in groups, more traditionally masculine men are more drawn to competitive type social groups, more traditionally masculine men tend to have a more avoidant attachment to groups, and masculinity endorsement doesn&rsquo;t affect men&rsquo;s level of participation in groups. Keywords: men, attachment, masculinity, groups, gender, norms</p><p>

The increasing number of young people who are inactive and not engaged in education, employment or training (NEETs) in the UK over the last years bears severe implications both for individual young people and for the society. This study explores the processes underlying the effects of neighborhood context on young people who experience NEET status. It relies on quantitative data from a nationally representative study, the Longitudinal Study of Young People in England (LSYPE), linked with the seven decomposed English Indices of Deprivation. Drawing on previous sociological theories this study puts forward an original theoretical framework, the Ecological Model of Neighbourhood Effects that proposes four pathways that mediate the direct effect of neighbourhoods on young people: a) individual characteristics and attitudes; b) parental characteristics and relationships; c) school experiences and attitudes to schooling, and; d) social epidemics. Potential causal pathways between neighbourhood context and individual outcomes are explored on a first strand of analysis by employing a logistic regression model. The results show that there is a higher probability for young people who live in high Crime Score areas to become NEETs in comparison to those who live in areas with low Crime Score after controlling for individual, family, school and peer group characteristics. On a second strand of analysis, I employ counterfactual models, propensity score matching and sensitivity analysis. The findings suggest that when two groups of children with identical observed characteristics at the age 13/14 experience di↵erent neighbourhood contexts, those who grow up in high Crime Score areas are more likely to become NEETs in comparison to those who grow up in low Crime Score areas. Unobserved characteristics though indicate the presence of selection bias that could alter the inferences drawn about neighbourhood effects.

<p> This study aims to advance understanding of the impact of higher education on students&rsquo; career development in today&rsquo;s more diverse student body, and to revisit the career development model to meet the evolution of the labor market. A latent structure incorporating three indicators is developed to measure college-prepared career readiness, i.e. generic competencies, subject-specific skills, and degree completion status. Using a national dataset of the 2004/09 Beginning Postsecondary Students Longitudinal Study (BPS: 04/09), this dissertation examines college students&rsquo; career readiness and the impact on their labor market outcomes in terms of job status and income six years after entering college. The central research questions are: 1) what is the status of students&rsquo; career readiness acquired through higher education, and what are the individual and institutional factors that affect college-prepared career readiness; 2) what are the relationships between college-prepared career readiness and labor market outcomes, and what are the student-level and institution-level variables associated with students&rsquo; labor market outcomes. A series of two-level hierarchical linear regression models are developed to address the research questions. </p><p> Results suggest that American students&rsquo; college-prepared career readiness is associated with multiple individual factors, including gender, ethnicity and immigrant status, academic integration, majoring STEM fields, high school GPA and college entrance exam scores. At the institution-level, selectivity and school control are related to college-prepared career readiness. With respect to labor market outcomes, college-prepared career readiness plays the most important role among all variables involved in predicting students&rsquo; job status and job income, at both student- and institution-level. In addition, socio-economic status, college major, duration of employment, academic integration and college entrance exam scores are variables affecting job status at the student-level. For job income, gaps favoring male and students majoring STEM fields are found at the student-level. Mean SES and research emphasis are associated with job income at the institution-level. </p><p> The findings provide evidence for university-wide programs designed to boost students&rsquo; academic integration, and partnerships between STEM departments and predominantly minority or high-need school districts. Particularly, there is a need for special career services that assist females in STEM fields access to equal opportunities.</p><p>

The objective of this research paper was to examine the possibility of active investment out-performing the passive investment by using a combination-based investment style for an extensive period. The combination-based style included financial-ratio-based style, market-based style and behavioural-finance-based style in the Johannesburg Stock Exchange during the period from 1984 to 2014. The four-dimension optimisation exercise based on the combination-based style was done in the in-sampling period and the result was tested in the out-of-sample period. The results have confirmed that the combination-based style out-performed the benchmark by 13% per annum over a 14 year period, which suggested that active managers could out-perform passive investment. The out-performance could further improve by recalibrating the optimisation exercise throughout the out-of-sample period to ensure the investment style learns from and incorporates with new data. / Dissertation (MBA)--University of Pretoria, 2014. / zkgibs2015 / Gordon Institute of Business Science (GIBS) / MBA / Unrestricted