Since 1949 Taiwan and China have been governed by different political regimes. Nevertheless, after more than fifty years, research shows that women in both societies now enjoy significantly higher social status and have lower fertility rates. Despite well-documented literature on the relationship between fertility and women??s status in Taiwan and China, no systematic empirical research has compared the two. This dissertation was designed to investigate the effects of women??s status on fertility and sterilization behaviors in China and Taiwan by means of multi-level analysis focusing on women??s education levels and employment status as predictors at both the individual and aggregate levels. To examine the influence of enforced policy, in China??s models, variables were added about whether the participants had a government-issued one-child certificate or had complied with the childbirth quota set by local authorities. Most results are consistent with our hypotheses. At the macro level, female college graduation rate is significant in Taiwan but not in China. One-child certificate rate is significantly correlated with provincial-level number of Child Ever Born (CEB). At the micro level of Poisson and logistic models, women with status are significantly more likely to have smaller numbers of CEB and lower sterilization usage. Survival analysis that simultaneously analyzed time duration and event occurrence showed dynamic effects of women??s status on the probability of a first, second and third childbirth. The Hierarchical Generalized Linear Models (HGLM) method shows both some direct and some interactive effects of contextual variables on fertility and contraceptive behaviors. In both countries, wives?? educational levels showed the greatest numbers of significant correlations with the dependent variables. Both Western socioeconomically based demographic transition theory and Asian planned demographic transition theory in China receive empirical support in the findings. Methodological and policy implications for future studies are discussed. The findings of this dissertation, particularly the micro-macro linkages, contribute to an explanation of how higher women??s status and lower fertility rates across the two regimes emerged from both common and disparate processes. This dissertation also illustrates how multi-level investigations of fertility and women??s status could be implemented in other parts of the world.

The primary objective of this research was to investigate the propensity to migrate the destination choices of young adults, and the importance of individual, household, and community characteristics in these migration choices. Using cohort data from the National Longitudinal Survey ofYouth79 from 1980 to 1998, this study specifies the set of individual-, household-, and community-level of determinants on migrat ion and then incorporates these variables in multivariate analyses to test their direct and relative effects on the migratory behavior of young adult groups. A Cox proportional hazard analysis suggests that among three levels of factors, individual characteristics are the most important determinants of migration, but the migratory behavior is more fully explained by multilevel variables rather than a single-level variable. This research had three foci within the primary objective. First, at the individual level, this study is tbe first step in research that intended to suggest the usefulness of status inconsistency arguments on migration studies. Findings of tbe research indicate that underrewarded individuals are more likely to migrate than those who have balanced status, while overrewarded individuals are less likely to migrate than those who have balanced status. Second, at the household-level investigation, this research focused on the effects of relative conjugal power between husbands and wives on migration. Results suggest that differences in relative power between husbands and wives has only minor effects on migration and the direction of migration, but the quantitative effects of relative power variables are greater for wives than for husbands. Third, at the community-level investigation, this study focused on analyzing the interaction between the residential mobility of individuals and characteristics of the residential areas where they are located. The migration propensity of the most mobile types of people (the more educated whites) has responded more to differences in community characteristics than that of the least mobile types of people (the less educated blacks).

Increasing shares of European women are making large investments in their human capital. Whether and to what extent these investments are in conflict with reproductive behaviour are issues that have repercussions for fertility levels. Using two Eurobarometer survey data (2006 and 2011) on individuals clustered in the 27 EU countries, I investigate the relationship between women's education and lifetime fertility intentions. Results suggest that a positive association between women's level of education and lifetime fertility intentions exists at both the individual and country levels, as well as in a micro-macro integrated framework. The main explanation for these findings - which remains to be proven by future research - is that, in institutional contexts allowing highly educated women to have large families, women of reproductive ages are more prone to make investments in both human capital and family size, because these choices are not seen as incompatible alternatives. (author's abstract)

In modern China, air pollution has become an essential environmental problem. Over the last 2 years, the air pollution problem, as measured by PM 2.5 (particulate matter) is getting worse. My report aims to carry out a longitudinal data analysis of the air quality index (AQI) in 3 main economic zones in China. Longitudinal data, or repeated measures data, can be viewed as multilevel data with repeated measurements nested within individuals. I arrive at some conclusions about why the 3 areas have different AQI, mainly attributed to factors like population, GDP, temperature, humidity, and other factors like whether the area is inland or by the sea. The residual variance is partitioned into a between-zone component (the variance of the zone-level residuals) and a within-zone component (the variance of the city-level residuals). The zone residuals represent unobserved zone characteristics that affect AQI. In this report, the model building is mainly according to the sequence described by West et al (2007) with respect to the bottom-up procedures and the reference by Singer, J. D., & Willett, J. B (2003) which includes the non-linear situations. This report also compares the quartic curve model with piecewise growth model with respect to this data. The final model I reached is a piece wise model with time-level and zone-level predictors and also with temperature by time interactions. / text

The purpose of this study was to begin testing of a downward cross-level model for studying the ability of older adults to transition from a nursing home after a Medicare Part A reimbursed stay. Transitions are known to be a weak point in the provision of healthcare to older adults and thus far, research has not identified those factors that influence older adult's transitions i.e., from the nursing home after a post acute stay. The theoretical background for this study was supported by Resource Dependency Theory which is a theory that contends that organizations are externally controlled by activities outside the organization such as the "free-market" economic model that predominates the nursing home industry. It was thought that nursing homes may prioritize their need for resident census above the resident's need for discharge. The hypothesis was that both individual resident characteristics and organizational characteristics might influence the ability of older adults to transfer from the nursing home after a Medicare Part A stay. The method of analysis in this study was contextual regression. Individual and facility characteristics were the independent variables and length of stay was the dependent variable. For this project, emphasis was placed on the development of a methodology for using the MDS in this and future research studies. Selection of variables and methods for variable computation were highlighted. Individual and facility characteristics and discharge disposition (level of care) were reported descriptively. Although facility characteristics did not contribute significantly to the model, individual characteristics explained 28% of the variance in the length of stay. Fifteen percent of individuals in the sample died during their Medicare Part A stay and 18% were readmitted to the hospital. The most prevalent diagnoses of the sample were hypertension (35%), falls (34%) and arthritis (32%). Findings suggest that individual characteristics account for only a portion of the length of stay for post acute nursing home residents. Further model testing is needed and should include a larger facility sample size and market characteristics to determine if those factors significantly influence the ability of older adults to transfer after the Medicare Part A stay ends.

Black students’ performance in mathematics on standardized examinations compared to White students is dismal; however, previous research has shown that there are highly effective high schools for Black students as defined by high levels of mathematics performance. Underachievement in upper-level high-school mathematics courses is a barrier for Black students’ access to many postsecondary education opportunities and contributes to an underrepresentation of Black students in the science, technology, engineering, and mathematics college majors and related careers. This dissertation examines the multilevel characteristics (student, teacher, department, and school factors) of high school mathematics programs as measured by Black students’ performance on standardized tests in 12th grade. The data for this study were taken from the Education Longitudinal Study (ELS) of 2002 from the National Center for Education Statistics of the U.S. Department of Education. Results indicated that the 12th grade achievement in mathematics of Black students is positively related to prior mathematics achievement, family SES, and the interaction between the teaching practices of mathematics teachers and students’ collective sense of safety. A surprising interaction effect indicated that in schools where there is collective sense about lack of safety and where students report low-quality mathematics teaching practices, Blacks are more likely to exhibit high student-achievement in mathematics than schools where the collective sense about safety is mixed (or where it is high) or where there is student consensus of good quality teaching practices.

Background: Symptomatic heart failure is a chronic and disabling condition that affects over 350 000 Canadians and is characterized by inevitable progression. Historically, research on the ways to increase survival has focused on biomedical factors. However, the continued poor prognosis of heart failure has prompted the search for other ways to improve the lives of these patients. Research in other chronic conditions demonstrates that social circumstances, described collectively as individual social interactions (e.g. social support, social participation) and community social factors (e.g. social capital, social norms), can influence health outcomes. Purpose: The purpose of this research was to describe and assess the impact of selected social circumstances potentially related to heart failure outcomes. Methods: Two literature reviews and one empirical study were performed. Conceptual models were proposed to describe the hypothesized pathways between selected social circumstances and heart failure outcomes. The first review was a systematic review of quantitative studies evaluating the relationship between social support and both rehospitalization and death. The review included a critical analysis of the methods employed by previous studies. The second review integrated the qualitative and quantitative literature describing the relationship between individual social interactions (including support, roles and participation) and the quality of life of patients or experience of living with heart failure. A narrative summary was provided and an integration of findings from both qualitative and quantitative study designs was performed. In the empirical study, patients‘ demographic and clinical information was examined simultaneously with selected community factors in a multilevel analysis. Outcomes of interest included rehospitalization or death of heart failure patients. Results: The systematic review shows that previous quantitative research has linked social support to reduced rehospitalization, but there is little evidence to link it with prolonged survival. The critique of the methods describes an inadequate conceptualization and inconsistent measurement of social support. A conceptual model showing how social support can influence rehospitalization is proposed. The integrative review presents qualitative research that identified the following social interactions as important components of the heart failure experience: social support, social participation and role fulfillment. However, no quantitative relationship between social support and quality of life was found. The potential reasons for the discrepant findings between the qualitative and quantitative studies include: the focus on social support as the only component of social interactions assessed in the quantitative literature; and the inconsistent measurement of social support. A conceptual model is presented to describe the multiple components of social interactions and the theoretical basis for their effects. The multilevel analysis demonstrates that individual factors exerted the strongest effect on heart failure outcomes in most models. Community characteristics had little influence on rehospitalization or death. Study design and analysis issues are proposed to explain these findings. Conclusion: The literature reviews and the empirical study provide a contribution to the population health literature, offering a broad approach to assessing the determinants of disease progression in heart failure patients. This thesis research advances the discussion about which social circumstances may influence heart failure outcomes and their pathways. The use of the proposed conceptual models in future research will help clarify the role of social circumstances in the prognosis of heart failure.

<p>Background: Childrens health, particularly in the early years, forms the basis of future health and development and plays a significant role in predicting individual life and opportunities. Thus, studies which enhance the understanding of the determinants of childrens health status are needed. Previous research on childrens health had focused on the familys and infants characteristics and ignored the potential impact of macro-level influences. The objectives of this thesis were (i) to examine the independent effects of neighbourhood factors on childhood health outcomes, (ii) to explore neighbourhood moderating effects on the associations between some individual risk factors and childhood health outcomes, and (iii) to quantify the contribution of neighbourhood factors to childhood health outcomes.</p><p>Method: The study population included 9,888 children born to women residing in Saskatoon during three years, 1992-1994. The data used in this study were extracted from three sources. The information related to birth outcomes and the mothers characteristics was extracted from the birth registration files maintained by Saskatchewans Vital Statistics Branch. The health services utilization information was generated from Saskatchewan Healths computerized administrative databases. The information related to the neighbourhood characteristics was obtained from Statistics Canadas 1991 Census, from local sources such as the Planning Department of the City, and two specialized neighbourhood surveys. Six domains of neighbourhood were examined in this study: socio-economic disadvantage, social interaction, physical condition, population density, local programs and services, and unhealthy lifestyle norm. This study was divided into two focused topics corresponding to two childrens health outcomes: low birth weight (LBW) and childrens hospitalizations (both incidence and length of stay). Multilevel modelling was employed to examine the independent/moderating impacts of neighbourhood characteristics on these childrens health outcomes. GIS mapping was used to visualize the associations between neighbourhood characteristics and childrens health outcomes.</p><p>Findings of focused topic 1: There was a significant variation across Saskatoon neighbourhoods in the distribution of LBW rate. This significant variation was attributed to both the characteristics of individuals living within the neighbourhoods as well as the characteristics of the neighbourhood of residence. Neighbourhood variables were both independent risk factors for LBW and moderators for the association between maternal characteristics and LBW. Specifically, a greater level of socio-economic disadvantage, a lower level of program availability and accessibility within the neighbourhoods were associated with a higher risk of LBW. A significant interaction between neighbourhood social interaction and single parent status was found. The risk of single parent status on LBW was mitigated by a greater level of social interaction within neighbourhoods. With individual level variables held constant, three neighbourhood variables predicted LBW, together contributing to a change in LBW rate of 7.0%.</p><p>Findings of focused topic 2: This focused topic employed a longitudinal/multilevel design to examine the effects of socio-economic status at multiple levels on childrens hospitalization. The key findings of this focused topic are the following: (i) There was a gradient association between the number of adverse birth outcomes and childhood hospitalization; (ii) There was a significant interaction between family income and adverse birth outcomes (i.e., the effect of adverse birth outcomes on childhood hospitalization was heightened among those children living in low income families); (iii) Neighbourhood characteristics, specifically neighbourhood socio-economic disadvantage, neighbourhood physical condition, and neighbourhood population density had independent effects on childhood hospitalization over and above the effect of family income; (iv) With individual level variables held constant, three neighbourhood variables (i.e., neighbourhood socio-economic disadvantage, physical condition and population density) together accounted for a variation of 40% in the incidence rate of hospitalization, and two neighbourhood variables (i.e., neighbourhood socio-economic disadvantage and physical condition) together accounted for a change in the length of stay per hospitalization from 2.88 days to 5.18 days across neighbourhoods.</p><p>Conclusion: Both individual and neighbourhood characteristics determined childhood health outcomes examined. Neighbourhood factors acted as independent risk factors as well as moderators on the association between individual risk factors and health outcomes. The contribution of neighbourhood factors to childrens health outcomes was quite substantial. The findings suggest that future interventions aimed at improving childrens health status in Saskatoon may be enhanced by targeting both high risk individuals and high risk neighbourhoods. The geographical variations in childrens health outcomes reported in this study are modifiable; they can be altered through public policy and urban planning, and through the efforts of families and children.

Background: Symptomatic heart failure is a chronic and disabling condition that affects over 350 000 Canadians and is characterized by inevitable progression. Historically, research on the ways to increase survival has focused on biomedical factors. However, the continued poor prognosis of heart failure has prompted the search for other ways to improve the lives of these patients. Research in other chronic conditions demonstrates that social circumstances, described collectively as individual social interactions (e.g. social support, social participation) and community social factors (e.g. social capital, social norms), can influence health outcomes. Purpose: The purpose of this research was to describe and assess the impact of selected social circumstances potentially related to heart failure outcomes. Methods: Two literature reviews and one empirical study were performed. Conceptual models were proposed to describe the hypothesized pathways between selected social circumstances and heart failure outcomes. The first review was a systematic review of quantitative studies evaluating the relationship between social support and both rehospitalization and death. The review included a critical analysis of the methods employed by previous studies. The second review integrated the qualitative and quantitative literature describing the relationship between individual social interactions (including support, roles and participation) and the quality of life of patients or experience of living with heart failure. A narrative summary was provided and an integration of findings from both qualitative and quantitative study designs was performed. In the empirical study, patients‘ demographic and clinical information was examined simultaneously with selected community factors in a multilevel analysis. Outcomes of interest included rehospitalization or death of heart failure patients. Results: The systematic review shows that previous quantitative research has linked social support to reduced rehospitalization, but there is little evidence to link it with prolonged survival. The critique of the methods describes an inadequate conceptualization and inconsistent measurement of social support. A conceptual model showing how social support can influence rehospitalization is proposed. The integrative review presents qualitative research that identified the following social interactions as important components of the heart failure experience: social support, social participation and role fulfillment. However, no quantitative relationship between social support and quality of life was found. The potential reasons for the discrepant findings between the qualitative and quantitative studies include: the focus on social support as the only component of social interactions assessed in the quantitative literature; and the inconsistent measurement of social support. A conceptual model is presented to describe the multiple components of social interactions and the theoretical basis for their effects. The multilevel analysis demonstrates that individual factors exerted the strongest effect on heart failure outcomes in most models. Community characteristics had little influence on rehospitalization or death. Study design and analysis issues are proposed to explain these findings. Conclusion: The literature reviews and the empirical study provide a contribution to the population health literature, offering a broad approach to assessing the determinants of disease progression in heart failure patients. This thesis research advances the discussion about which social circumstances may influence heart failure outcomes and their pathways. The use of the proposed conceptual models in future research will help clarify the role of social circumstances in the prognosis of heart failure.

The main goal of this study is to investigate the spatial patterns of police-reported crime rates across select Canadian urban neighbourhoods and to explore their relationships with both neighbourhood- and city-level characteristics, as well as neighbourhood spatial dependence. Analyses were based on aggregated data from the 2001 Incident-Based Uniform Crime Reporting Survey (UCR2) and the Census of Population for six Canadian cities: Edmonton, Halifax, Montreal, Saskatoon, Thunder Bay and Toronto. Exploratory spatial data analysis (ESDA) was used to examine the spatial distribution of crime as well as to test for spatial dependence in the crime data. By using multilevel modelling and spatial regression techniques, neighbourhood violent and property crime rates were modeled respectively as a function of both city- and neighbourhood-level contextual variables while controlling for spatial dependence. The results show that crime is not distributed randomly, but tends to be concentrated in particular neighbourhoods, notably around the city centers of these cities. Neighbourhood variance in crime rates is not only dependent on local neighbourhood characteristics, but also on the characteristics of surrounding neighbourhoods, as well as the broader city environment where neighbourhoods are embedded. These findings suggest that strategies aimed at preventing or reducing crime should be developed in light of specific local neighbourhood contexts, while taking into account social forces external to the immediate neighbourhood in the wider social environment.