Latent variable models (LVMS) are widely appreciated multivariate methods to explore variables that are related to the observed variables, and assessing the relationships among them. One of most widely used latent variable models is structural equation model (SEM). Based on more than a dozen standard packages for fitting SEMs, such as LISREL VIII (Jorskog and Sorbom, 1996), and EQS (Bentler, 2004), these models have been widely appreciated in behavioral, educational, medical, social, and psychological research. The statistical theories and methods in these packages are based on the normal distribution; hence, they are vulnerable to outliers and the non-normal assumption. As outliers and non-normal data set are commonly encountered in substantive research, this fundamental problem has received much attention in the field. However, almost all existing methods are developed via the covariance structure analysis approach that heavily depends on the asymptotical properties of the sample covariance matrices S. Hence, this approach cannot be applied to the more complex SEMs and/or SEMs with more complex data structure such as missing data, because under these more complicated situations S is complicated, and its asymptotical properties are not well known. The objectives of this thesis are to develop novel robust methods for analyzing complex SEMs and/or more data structures, including but not limited to nonlinear SEMs with missing data. Both maximum likelihood (ML) and Bayesian approaches for estimation, hypothesis testing and model comparison will be investigated. Efficient algorithm for computing the results for statistical inference will be developed through unitization and modification of the advanced tools in statistical computing, for example the Monte Carlo Expectation-Maximization algorithm, and the Markov Chains Monte Carlo methods. Asymptotical properties of some statistics are derived. Simulation studies and real examples are conducted to reveal the empirical performance of the Bayesian and ML approaches. The newly developed methodologies will be very useful for analyzing complex data in the substantive research. / Xia Yemao. / "October 2005." / Adviser: S. Y. Lee. / Source: Dissertation Abstracts International, Volume: 67-07, Section: B, page: 3883. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (p. 105-114). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract in English and Chinese. / School code: 1307.

The business network that is linked to flexibility, aggressiveness and strategy has become increasingly important in recent years. Several studies suggest that such networks potentially have a profound impact on firm performance, including Small and Medium-sized Enterprises (SMEs). The role of SMEs in enhancing global and local economic growth is undeniable, particularly in the context of developing countries such as Malaysia. Although numerous researches have been conducted in this field, the majority of them limit their focus to the relationship between firm capabilities (i.e. innovation and dynamic capabilities) and firm performance in specific industries. Research on the synergy impact of business networks, innovation and dynamic capabilities on SME performance remains scarce. This has become a significant gap, which this research seeks to address. This research investigates the roles of dynamic capabilities and innovation capabilities as a moderator and mediator in the relationship between business networks and firm performance, based on the model developed from the concepts of the Resource Base View (RBV) and Dynamic Capability (DC) theories. The model was justified through the Structural Equation Modelling (SEM) technique using AMOS version 23. Taking Malaysia as a research context, the model was tested against a total of 463 SMEs across different industries and categories (i.e. micro, small, and medium SMEs) through face-to-face surveys with 130 owners, 41 CEOs, 79 managers and 213 executives. This study presents five important findings: (1) there exists no direct relationship between business network and firm performance; (2) there exists a direct relationship between innovation, dynamic capabilities and firm performance; (3) the existence of the relationship between business network and firm performance is conditioned by innovation capabilities; (4) dynamic capabilities do not moderate the relationship between business network and firm performance; however (5) dynamic capabilities moderate the relationship between business networks and innovation capabilities. To conclude, the synergy of business networks, innovation capabilities and dynamic capabilities will significantly affect SME performance. This implies that SME performance will not be affected by the business network, as a single variable. The research offers three key contributions. Firstly, it enhances our understanding of the important synergies between business networks, innovation capabilities and dynamic capabilities in elevating SME firm performance. Second, the findings provide a new perspective on how the application of RBV and DC theories can be used as a conceptual lens to analyse the factors affecting SME performance. Lastly, the result signposts practical approaches for SME decision-makers by providing assistance to boost firm performance.

Levels of non-resident parenthood in Scotland are substantial. The 2011 Scottish Census indicated 28 per cent of all families with dependent children to be lone parent households. Whilst non-resident parenthood is not synonymous with non-resident fatherhood, 92 per cent of such households were headed by the mother (ONS, 2014). Child well-being in non-resident father households is an issue of concern amongst policy makers and practitioners and both law and policy appear to operate on the principle that the maintenance of non-resident father-child relationships is generally conducive to child well-being. Whilst there is evidence to suggest the well-being of children in non-resident father households is typically poorer than their contemporaries in two natural parent households (Amato and Keith, 1991; Amato, 2005), and indeed evidence to suggest non-resident father involvement may benefit child well-being (Amato and Gilbreth, 1999; Adamsons and Johnson, 2013), the existing research has a number of limitations. Firstly, studies have typically adopted narrow conceptualisations of child well-being. Secondly, few studies have sought to disaggregate the total effects of non-resident fatherhood to consider both those transmitted directly and indirectly via mediating variables. Finally, increasing numbers of non-marital births coupled with evidence suggesting cohabiting relationships to be at an increased risk of breakdown in children’s early years compared to marriages (Greaves and Goodman, 2010), has culminated in increasing levels of early years non-resident fatherhood, an issue which has received less attention in the literature. Using data from the Growing Up in Scotland study this research explores associations between non-resident fatherhood and child well-being and the potential pathways through which such associations may operate. The research conceptualises child well-being as a multi-dimensional construct comprising four key domains: social, emotional and behavioural development, cognitive development, general health and material resources and uses structural equation modelling to consider the extent to which firstly, living in a non-resident father household, and secondly, non-resident father involvement is associated with child well-being directly, or indirectly, via household income, maternal mental health and parenting behaviours. The results show that, relative to two natural parent households, child well-being across each of the four domains is poorer in non-resident father households headed by a lone mother but not in those where the mother has re-partnered. Only one statistically significant direct association was found in the domain of material resources with the results suggesting much of the negative association to be transmitted indirectly via household income and maternal mental health. For those children who were reported as having some form of contact with their father, the results indicate paternal involvement to be indirectly associated with fewer social, emotional and behavioural difficulties via maternal mental health. Finally, consideration of the circumstances and characteristics associated with the maintenance of contact and levels of paternal involvement revealed maternal relationship status to be an important correlate of both contact and involvement whilst parental relationship history and circumstances surrounding the pregnancy were additionally found to be important correlates of contact. This thesis argues that simple dichotomies of father presence / absence can serve to mask the complex network of relationships underlying associations between non-resident fatherhood and child well-being. It also argues that child well-being in non-resident father households would benefit from policies which seek to alleviate financial hardship and support maternal mental health. It suggests that the maintenance of non-resident father-child relationships is potentially beneficial for child well-being and argues that promotion and support of contact and involvement in children’s early years likely requires a targeted approach. It further argues that the role of the non-resident father should be construed broadly to include consideration of support for the wider household in which the child lives.

O preço de um imóvel é definido como um preço hedônico, visto que sua formação é dada através de um conjunto de atributos que consideram, além das características do próprio bem, tais como número de quartos, vagas na garagem, área privativa, entre outros; as características do bairro onde este imóvel está inserido. Existe uma demanda crescente por imóveis localizados em regiões com uso de solo misto, densificadas e com fácil acesso a serviços. Essas características estão fortemente vinculadas ao conceito de caminhabilidade. Juntamente com facilidade de acesso ao transporte público, desenho urbano que favoreça o deslocamento dos pedestres e segurança pública, formam as 6 dimensões da caminhabilidade utilizadas nesse estudo. Bairros com essas características estimulam viagens a pé e estima-se que haja uma valorização imobiliária de seus empreendimentos, com consumidores dispostos a pagar mais por imóveis residenciais localizados em regiões caminháveis. Para testar o impacto da caminhabilidade no valor do preço, foram utilizados modelos de equações estruturais. Eles permitem que conceitos como os de caminhabilidade e segurança pública, que não podem ser diretamente medidos (variáveis latentes), possam ser explicados por medidas observáveis que atuam como formadoras desses conceitos. Os resultados obtidos confirmam a hipótese, com o preço do metro quadrado de imóveis residenciais a venda crescendo conforme aumenta a caminhabilidade. A segurança pública demonstrou ser o fator mais importante na explicação da caminhabilidade e consequentemente na valorização do preço. / Sale price of a property is defined as a hedonic price, since its value is formed by a set of attributes that cover more than just characteristics of the product itself, such as number of rooms, parking spots and private area. There is a growing demand for properties located in areas with mixed land use, densified and with easy access to products and services. These characteristics are strongly linked to the concept of walkability, usually named as walkability dimensions. Neighborhoods with these features encourage commutes by walking and are likely to have a real estate valuation, with consumers willing to pay more for residential properties located in walkable areas. Structural equation models were used to test the impact of walkability on sales price. This methodology allows to quantify concepts as walkability and security, that could not be directly measured (latent variables), by using the observable variables that compose them. The results confirm the hypothesis that increase in walkability will also raise the sales price. Security has proven to be the most important factor in the walkability explanation and consequently in price appreciation.

Residentes de um mesmo contexto urbano podem apresentar padrões distintos de deslocamento. A distinção desses padrões pode estar relacionada às diferenças na percepção acerca da caminhabilidade do entorno. Para estimular mudanças no padrão de deslocamento dos usuários de transporte motorizado é importante compreender como os fatores do ambiente urbano, percebidos por esses usuários, se diferenciam em relação aos usuários de transporte ativo, e a partir disso montar estratégias eficientes para cada público alvo. A partir de uma pesquisa realizada com moradores de bairros da zona leste do município de São Paulo, este trabalho compara os usuários de transporte ativo, coletivo e individual em relação a três questões: (i) os fatores que influenciam a percepção da caminhabilidade (dentre eles a segurança pública), (ii) o impacto da caminhabilidade na satisfação com o bairro e (iii) a relação entre percepção de caminhabilidade, satisfação com o bairro e o impacto nos hábitos de deslocamento ativos (frequência e tempos de viagem semanais) reportados para os três grupos. Para analisar tais relações, foram desenvolvidos modelos de equações estruturais, que permitem testar uma hipótese entre múltiplas variáveis e que diversas variáveis observadas representem conceitos que não podem ser diretamente medidos (variáveis latentes), como a percepção de caminhabilidade e a sensação de segurança pública Os resultados obtidos mostraram a existência de diferenças na percepção do ambiente construído entre usuários de transporte ativo, transporte coletivo e transporte individual motorizado. As diferenças principais foram encontradas entre os usuários de transporte ativo e os usuários de transporte motorizado (coletivo e individual), tanto em relação à percepção do ambiente, quanto em relação à sua influência no comportamento do usuário relativo à viagem. A preocupação com a segurança pública parece ser superestimada por aqueles que não realizam deslocamentos ativos regularmente. Investimentos em segurança pública e na qualidade do ambiente mostram ter um efeito positivo em todos os usuários, tanto no que se refere à satisfação com o bairro quanto à utilização de modos ativos de deslocamento. O impacto da percepção do ambiente na satisfação com o bairro foi maior para os usuários de transporte coletivo e individual. / Residents of a same urban context may present different travel patterns. The distinction of these travel patterns may be related to differences in the perception of the environment walkability. To stimulate changes in the travel pattern of motorized transport users, it is important to understand how the environmental factors perceived by these users differ in relation to the users of active transport, and from that, create efficient strategies for each target public. Based on a survey with residents of neighborhoods in the eastern part of the city of São Paulo, this study compares users of active, public and private transport in relation to three issues: (i) the factors that influence perceived walkability (among them public security) (ii) the impact of walkability on neighborhood satisfaction, and (iii) the relationship between perceived walkability, neighborhood satisfaction, and the impact on active commuting patterns (weekly frequency and travel times) reported for all three groups. To analyze such relationships, structural equation models have been developed, which allow testing a hypothesis among multiple variables and that several observed variables represent concepts that cannot be directly measured (latent variables), such as perceived walkability and sense of public security The results showed the existence of differences in the perception of the built environment between users of active, public and private transport. The main differences were found between users of active and users of motorized transport (public and private), both in relation to the perception of the environment and its influence on travel user behavior. Concern about public safety seems to be overestimated by those who do not regularly commute by active means. Investments in public security and environmental quality have a positive effect on all users, both regarding neighborhood satisfaction and the use of active modes of travel. The impact of perceived environment on neighborhood satisfaction was greater for users of public and individual transport.

李晓煦. / Submitted: Jan. 2010. / Thesis (doctoral)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (p. 170-175). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts in Chinese and English. / Li Xiaoxu.

Lam, Kwok Hap. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2009. / Includes bibliographical references (leaves 68-72). / Abstract also in Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Two-level NSEM with outcomes from Exponential Family --- p.6 / Chapter 2.1 --- Basic Model Description --- p.7 / Chapter 2.2 --- Generalization from Normal Distribution to Exponential Family Distributions --- p.9 / Chapter 2.3 --- Bayesian Analysis of the Model --- p.10 / Chapter 2.3.1 --- Posterior Analysis and Gibbs Sampler --- p.10 / Chapter 2.3.2 --- Prior Distributions --- p.11 / Chapter 2.3.3 --- Bayesian Estimation --- p.13 / Chapter 2.3.4 --- Bayesian Model Selection --- p.14 / Chapter 2.4 --- A Simulation Study --- p.15 / Chapter 3 --- Two-level NSEM with mixed continuous and ordered categorical data --- p.28 / Chapter 3.1 --- Model Description --- p.29 / Chapter 3.2 --- Bayesian Analysis of the Model --- p.30 / Chapter 3.2.1 --- Posterior Analysis and Gibbs Sampler --- p.30 / Chapter 3.2.2 --- Bayesian Estimation --- p.31 / Chapter 3.3 --- A Simulation Study --- p.31 / Chapter 4 --- "Two-level NSEM with mixed continuous, count and binomial data" --- p.36 / Chapter 4.1 --- Model Description --- p.37 / Chapter 4.2 --- Bayesian Estimation --- p.38 / Chapter 4.3 --- A Simulation Study --- p.39 / Chapter 5 --- Two-level NSEM with mixed continuous and unordered categorical data --- p.43 / Chapter 5.1 --- Basic Model Description --- p.44 / Chapter 5.2 --- Bayesian Analysis of the Model --- p.47 / Chapter 5.2.1 --- Posterior Analysis and Gibbs Sampler --- p.47 / Chapter 5.2.2 --- Prior Distributions --- p.48 / Chapter 5.3 --- A Simulation Study --- p.49 / Chapter 6 --- Conclusion and Discussion --- p.53 / Chapter A --- Technical Details for Chapter 2 --- p.56 / Chapter A.1 --- Full conditional distributions --- p.56 / Chapter A.2 --- Implementation of the Metropolis-Hastings (MH) Algorithm --- p.59 / Chapter A.3 --- Gelman-Rubin statistic --- p.61 / Chapter B --- Technical Details for Chapter 3 --- p.63 / Chapter B.1 --- Full conditional distributions --- p.63 / Chapter B.2 --- Implementation of the Metropolis-Hastings (MH) Algorithm --- p.64 / Chapter C --- Technical Details for Chapter 5 --- p.66 / Chapter C.l --- Full conditional distributions --- p.66 / Bibliography --- p.68

Wang, Chang. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2008. / Includes bibliographical references (leaves 139-145). / Abstracts in English and Chinese. / ABSTRACT --- p.3 / CHINESE ABSTRACT --- p.5 / ACKNOWLEDGMENTS --- p.6 / TABLE OF CONTENTS --- p.7 / LIST OF TABLES --- p.9 / LIST OF FIGURES --- p.10 / INTRODUCTION --- p.11 / Chapter 1.1 --- ISSUE OF MODEL FIT IN SEM --- p.11 / Chapter 1.2 --- CLASSIFICATION AND DEVELOPMENT OF FIT INDICES --- p.13 / Chapter 1.3 --- ORGANIZATION OF THIS THESIS --- p.18 / Chapter CHAPTER 2 --- ISSUES OF FIT INDICES IN ASSESSING MODEL FIT --- p.19 / Chapter 2.1 --- SENSITIVITY OF FIS TO MODL PARAMETER --- p.19 / Chapter 2.1.1 --- Sample size --- p.20 / Chapter 2.1.2 --- Model complexity --- p.21 / Chapter 2.1.3 --- Misspecification --- p.23 / Chapter 2.2 --- MEASUREMENT ERROR --- p.26 / Chapter 2.3 --- PERFECT FIT VS. APPROXIMATE FIT --- p.26 / Chapter 2.4 --- Minimum Fit Function chi-square vs. Normal-theory Weighted Least chi-square --- p.29 / Chapter 2.5 --- RULE OF THUMB --- p.30 / Chapter 2.6 --- FIVE RESEARCH QUESTIONS --- p.37 / Chapter CHAPTER 3 --- SIMULATION --- p.39 / Chapter 3.1 --- FIT INDICES --- p.39 / Chapter 3.2 --- DESIGN OF MONTE CARLO SIMULATIONS --- p.38 / Chapter 3.3 --- MODEL COMPLEXITY AND MODEL SPECIFICATION --- p.39 / Chapter 3.4 --- SIMULATION PROCEDURE --- p.41 / Chapter CHAPTER 4 --- RESULTS --- p.45 / Chapter 4.1 --- MEASUREMENT ERROR AND CRONBACH´ةS ALPHA --- p.45 / Chapter 4.2 --- ANSWER TO Q1 --- p.45 / Chapter 4.3 --- ANSWER TO Q2 --- p.53 / Chapter 4.4 --- ANSWER TO Q3 --- p.56 / Chapter 4.5 --- ANSWER TO Q4 --- p.60 / Chapter 4.6 --- ANSWER TO Q5 --- p.62 / Chapter CHAPTER 5 --- DUSCUSSION --- p.77 / Chapter 5.1 --- DUSCUSSION TO Q1 --- p.77 / Chapter 5.2 --- DUSCUSSION TO Q2 --- p.83 / Chapter 5.3 --- DUSCUSSION TO Q3 --- p.85 / Chapter 5.4 --- DUSCUSSION TO Q4 --- p.88 / Chapter 5.5 --- DUSCUSSION TO Q5 --- p.89 / Chapter CHAPTER 6 --- LIMITATION --- p.99 / Chapter CHAPTER 7 --- CONCLUSION --- p.101 / PREFERENCE --- p.139

Pan, Junhao. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2009. / Includes bibliographical references (leaves 121-135). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

O ambiente competitivo e globalização vem provocando mudanças nas estratégias das empresas, fomentando, muitas vezes, para um processo de internacionalização. As empresas procuram internacionalizar-se de acordo com seus objetivos estratégicos, traçando uma estratégia de internacionalização. Sair do seu ambiente para um país desconhecido significa que a empresa estará sujeita a correr certos riscos e, portanto, requer uma estratégia que avalie os riscos envolvidos para se estabelecer em um determinado país. O conhecimento desses riscos vai proporcionar a empresa mitigar ou mesmo eliminar os riscos envolvidos. Independente do processo de entrada em outro país, sempre haverá a questão do risco. Geralmente, quando uma empresa decide se internacionalizar, ela opta primeiro pelo processo de exportação, em que terá de lidar com vários tipos de risco em suas decisões gerenciais para alcançar seus objetivos estratégicos. Portanto, essa pesquisa foi feita com oitenta e quatro empresas exportadoras brasileiras, do estado de São Paulo, com faturamento de até 10 milhões de dólares listadas na Confederação Nacional da Indústria (CNI). O objetivo desse estudo foi entender como e quais riscos envolvidos na internacionalização de empresas exportadoras brasileiras exercem influência sobre a estratégia e objetivos estratégicos na internacionalização de empresas exportadoras. Para a análise da pesquisa, foi utilizada a técnica de modelagem de equações estruturais baseada em variância, na qual foi utilizada uma survey para coleta de dados. Foram geradas quatro hipóteses, três relacionadas ao risco e estratégia e uma relacionada à estratégia e ao risco. Como resultado, duas hipóteses foram confirmadas e, portanto, duas rejeitadas. Essa pesquisa contribui com o conhecimento científico, pois, além de verificar os riscos que influenciam na estratégia e objetivos estratégicos, ela traz a concepção de envolver riscos não somente financeiros, mas também do ambiente, do setor e da empresa na estratégia para a internacionalização. / The competitive environment and globalization has been provoking changes in the strategies of companies, often leading to an internationalization process. The companies seek to internationalize themselves according to their strategic objectives, to which a strategy of internationalization will be traced. Leaving your environment to an unknown country means that the company is subject to certain risks and therefore requires a strategy that assesses the risks involved in establishing itself in a particular country. Knowing these risks will enable the company to mitigate or even eliminate the risks involved. Regardless of the process of entry into another country, there will always be the issue of risk. Usually, when a company decides to internationalize, it opts for the first time through the export process, where it will have to deal with various types of risk in its managerial decisions to reach its strategic objectives. Therefore, this research was conducted with eighty-four Brazilian companies in the state of São Paulo, with sales of up to 10 million dollars listed in Confederação Nacional da Indústria (CNI). The objective of the study was to understand how and what risks involved in the internationalization of Brazilian exporting companies influence the strategy and the strategic objectives in the internationalization of exporting companies. But even this process depending on the intensity of export the company will be subject to risk because it is unknown to the host country. Therefore, this research was carried out with Brazilian export companies, whose objective was to investigate the influence of the risks involved in the internationalization of Brazilian exporting companies on the strategy of internationalization and strategic objectives. For the analysis of the research was used the technique of modeling of structural equations based on variance, where a survey was used to collect data. Four hypotheses were generated, three related to risk and strategy and one related to strategy and risk. As a result, two hypotheses were confirmed and therefore two rejected. This research contributes with scientific knowledge, because in addition to verifying the risks that influence the strategy and strategic objectives, it brings the concept of involving not only financial risks but also the risk of the environment, sector and the company in the strategy for internationalization.