No description available.

No description available.

Team cohesion represents arguably the most studied team construct as it has been consistently shown to be associated with improved performance. However, although cohesion is now understood to be an emergent state—as it develops over a team's life cycle—research has yet to uncover the dynamic nature of cohesion. The current study was designed to particularly test the impact of team personality composition both on the initial status of cohesion, and on changes in cohesion over time. 80 newly formed teams performed a highly interdependent team task, and team cohesion was measured over six time points. Personality was measured prior to the task and calculated at the team level, as both an average and a variability score. After performing longitudinal hierarchical linear modeling, results indicated that team personality impacts cohesion differently at initial status and over time. In particular, higher team agreeableness predicted greater slopes of cohesion, but not initial cohesion levels. Also, higher extraversion predicted greater initial status of cohesion, but not greater slopes. These results present important boundary conditions for understanding the role of team personality composition on team cohesion. / Master of Science

For many years, researchers and educators have been concerned about achievement scores but seemed less interested in students’ happiness and psychological well-being at school. However, students’ psychological well-being experiences may facilitate students’ adjustment and ultimately lead to academic achievement. It can be assumed that students’ different motivational and well-being experiences in each course would contribute to students’ overall psychological well-being. The purpose of this study was to investigate how and why students experience different levels and kinds of motivation and well-being across courses. As the preliminary and important ground to allow me to address this purpose, I needed to establish first whether students experienced different levels of academic motivation and course well-being across the courses they were taking. A total of 505 students participated in this study and provided information about 1817 courses they were taking. The participants come from a subject pool of one department that attracted students from diverse majors. Multilevel modeling was used to explore different situational (Level 1) and personal experiences (Level 2) of motivation and course well-being across courses and across students. The unconditional model showed variability of perceptions at Levels 1 and 2 indicating that students did vary in their reports across courses and that nevertheless, there were individual differences across students in their aggregate experiences. The conditional model was used to test what course characteristics were associated with motivational and well-being indicators at the situational level. Course characteristics were taken from different constructs: course value, classroom structure, teacher characteristics, classroom goal structure, and a caring classroom climate. Predictors at the personal level included students’ sex and their perceptions of general needs for relatedness, general relatedness need fulfillment in everyday life, and personal growth. Having supported the preliminary hypothesis with the unconditional model that there was variance both within student and between students, I used the conditional model and found that various course characteristics were differently associated with academic motivation and course well-being. Overall, results addressed that teacher characteristics and a caring classroom climate were strongly associated with all the different kinds of motivational and course well-being indicators. Students’ personal characteristics were, also, differently related to these outcomes. / text

Propensity score matching is commonly used to estimate causal effects of treatments. However, when using data with a hierarchical structure, we need to take the multilevel nature of the data into account. In this thesis the estimation of propensity scores with multilevel models is presented to extend propensity score matching for use with multilevel data. A Monte Carlo simulation study is performed to evaluate several different estimators. It is shown that propensity score estimators ignoring the multilevel structure of the data are biased, while fixed effects models produce unbiased results. An empirical study of the causal effect of truancy on mathematical ability for Swedish 9th graders is also performed, where it is shown that truancy has a negative effect on mathematical ability.

Annually, nearly 7 million women and 5.5 million men experience some form of intimate partner violence, which has serious health impacts. IPV has also been shown to limit the impact of early childhood home visiting interventions. Given the positive impacts of home visiting, reducing IPV in that setting could alleviate the negative impacts of IPV and improve mother and child outcomes as well. The analysis performed are from data from a randomized trial of an intimate partner violence intervention that was embedded into the Nurse Family Partnership, an evidence-based home visiting program. The intervention focused on identifying IPV, and for women without severe IPV, improving relationship skills including communication patterns and conflict resolution. The goal of this analysis is to examine how changes in partner and history of IPV moderate the relationship skills outcomes. This study aims to fill the gap in knowledge regarding the relationship between a home visiting intervention and relationship skill outcomes in women enrolled in the home visiting program, the Nurse Family Partnership. The purpose of the study is to discover whether the relationship skills differ in participants with stable vs. unstable partnerships and with those who experienced IPV before the start of the study. Women were randomized to NFP as usual (n=105) or NFP+, which included NFP plus the IPV intervention (n=133). Participants were surveyed at baseline, and at one and two-year follow-up with 81% retention over 2 years. Standardized assessment tools assessed relationship quality, communication, problem solving, partner support, relationship decision making, and psychological maltreatment. Marginal modeling was conducted to examine whether the intervention accounted for any change in relationship variables and whether the impact is moderated by history of IPV and changes in partnership. Multilevel modeling of the outcome variables showed some main effects of time such that conflict resolution improves for the intervention group (p<0.05). There is one clinically significant three-way interaction showing reduced relationship danger in the intervention for women with a history of IPV (p<0.06). There were no significant interactions for the partnership change moderator.

Thesis advisor: John B. Williamson / Since the early 1980s, Uganda has been in the spotlight of global concerns about the HIV/AIDS epidemic that has almost brought the country to its knees. Consequently, a number of social epidemiologists and researchers from different social science fields have, over the past two and half decades, focused their attention on Uganda, attempting to identify the risk factors that expose people to HIV infection in order to inform intervention policy. Although studies coming out of this effort have provided important insights into risks of HIV infection, they have been criticized for almost entirely focusing on individual behavioral factors, such as prostitution and inconsistent condom use, as the primary causal factors of HIV infection, without comprehending the contextual background in which HIV infection takes place. Using the 2000/01 Uganda Demographic and Health Survey and employing multilevel logistic regression methods, I address this concern by investigating the influence of contextual factors on three behaviors related to the risk of HIV infection (HIV testing, multiple sexual partnering, and inconsistent condom use). Analyses reveal that educational attainment, socioeconomic status, and religion significantly predict HIV testing, multiple sexual partnering, and condom use for both men and women - and at both the individual and neighborhood levels. Analyses also reveal that age has an inverted U-shaped association with HIV testing and multiple sexual partnering for both men and women at the individual level. Despite important gains in slowing HIV infection rates over the past two decades, Uganda's increasing burden of the HIV/AIDS epidemic - amid faltering healthcare and other social services investments - is inevitable. It is apparent that there are formidable obstacles to effectively eradicating HIV/AIDS, unless essential social services - such as education, accessible healthcare services - are enhanced, and policies are introduced to improve socioeconomic status of individuals and entire neighborhoods. / Thesis (PhD) — Boston College, 2011. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Sociology.

Growth modeling in education has focused on student characteristics in multilevel growth accountability models and has rarely included financial variables. In this dissertation, relations of several demographic and financial characteristics of Oregon school districts to the reading and mathematics growth of students receiving special education services in Grades 3-8 were explored after accounting for student level demographic characteristics. Previous research indicated that three variables were potentially related to student growth: district level aggregated student demographics, district geography (e.g., location in a remote area), and district funding. Three sources of data were used to investigate these relationships: institutional data reported by the Oregon Department of Education, the Common Core of Data gathered by the National Center for Education Statistics, and Oregon Assessment of Knowledge and Skills test data collected as part of the National Center on Assessment and Accountability in Special Education. Multi-level models of student growth across Grades 3-8 were constructed for reading and mathematics, with time (level-1) nested within students (level-2) and districts (level-3). Results demonstrated that although student-level demographic factors account for the majority of meaningful differences in student growth, both district demographic characteristics and financial investment in students were related to growth for students who received special education services.

In the 1960s, the government formed the President's Commission on Law Enforcement and Administration of Justice to looked at the problem of crime and fear of crime in modern American society. In addition to looking at these issues, the Commission also looked at ways to potentially reduce both crime and fear of crime. One of the primary outcomes of the Commission's report was that policing agencies in the United States needed to fundamentally alter the way they served their communities, notably by transitioning to community-oriented policing (COP). Starting in the 1970s, law enforcement agencies around the nation began to embrace the COP philosophy in the hopes that it would effectively reduce crime. A plethora of research suggests that the crime reduction benefits of COP are dubious at best; however, COP shows great promise in reducing fear of crime in neighborhoods. However, scholars remain uncertain as to why COP can effectively reduce fear. The uncertainty surrounding the efficacy of COP lies in the incomplete theoretical understanding of fear of crime. Three largely divergent fear of crime models have been developed. The first, the social integration model, posits that fear is influenced by the degree to which a person is integrated into their community. The thought being that the more socially integrated a person is, the stronger the sense of informal social and thus the lower the fear of crime. Research generally--although not always--supports this notion. Other scholars developed the disorder model, which posits that disorderly conditions or other signs of incivility can lead residents to feel as though informal social control has broken down, and thus elevate levels of fear. Again, this notion is well supported in the research. The final model suggests fear of crime is a result of sociodemographic differences (e.g., gender and age) that make a person feel more vulnerable to victimization, and thus those feeling most vulnerable exhibit the highest levels of fear. The findings from this so-called vulnerabilities model receive inconsistent support in the research. The problem with the extant fear of crime research is that it largely relies on singular explanations of fear. In other words, it operates from the premise that one of the models described above is responsible for residents' levels of fear. Recently, scholars have begun developing multimodel explanations in an effort to improve criminologists' ability to explain fear of crime. However, this multimodel approach is not a complete theoretical model of fear because it fails to account for the likely existence of a reciprocal effect between fear of crime and social integration. Further, it fails to account for the effects of social context may exert on fear and the way in which neighborhood differences may condition the individual-level fear of crime relationships. This dissertation, using two data sources, attempts to predict fear of crime using a more complete fear of crime model than those used in much of the prior research. The first source of data used is the 2004 Hillsborough County Sheriff's Office community survey (N=1898), which was distributed to a random sample of households in unincorporated Hillsborough County. Additionally, to create measures of social context, this dissertation utilizes data from the 2000 United States Census for census designated places in unincorporated Hillsborough County--which serve as the proxy for neighborhoods (N=30). Based on theory and prior research, it was hypothesized that the best fear of crime model would contain elements from all three theoretical models developed in prior research. Additionally, it was hypothesized that there would be a significant and negative reciprocal effect from fear of crime to social integration. Finally, it was hypothesized that social context would condition the relationships between individual-level fear of crime predictors. As predicted by the hypothesis, the empirically strongest fear of crime model did contain elements from all three explanatory fear of crime models. Additionally as hypothesized, there was a significant reciprocal relationship between fear of crime and social integration. However, contrary to expectations the relationship was positive. In other words, fear of crime motivated residents to become more socially integrated in their neighborhoods. Finally, as hypothesized social context did condition the effects of the individual-level variables. However, contrary to the hypotheses proffered, social context augmented the size of the effect between the individual-level variables. The findings from this dissertation offer some interesting insights for scholars and posivy makers alike. The findings suggest that it is imperative to use a more complete (e.g., multimodel) approach when explaining fear of crime. Additionally, it is necessary to account for the reciprocal relationship between fear of crime and social integration; otherwise research will yield deceptive parameter estimates for social integration on fear of crime. Lastly, social context matters and needs to be considered in further research. However, the theoretical model in this dissertation--while a step forward--does not represent the theoretical model to explain fear of crime. The results suggest that the model may be even more complex than the model presented here. The results of this dissertation for policy makers suggest that community oriented policing strategies are likely an effective mechanism for reducing residents fear of crime for two reasons; 1) the strengthening of social integration programs in neighborhoods and 2) focusing on reducing disorder problems in neighborhoods. Study strengths and limitations, as well as directions for future research are discussed.

This dissertation focuses on issues related to fitting an optimal variance-covariance structure in multilevel linear modeling framework with two Monte Carlo simulation studies. In the first study, the author evaluated the performance of common fit statistics such as Likelihood Ratio Test (LRT), Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC) and a new proposed method, standardized root mean square residual (SRMR), for selecting the correct within-subject covariance structure. Results from the simulated data suggested SRMR had the best performance in selecting the optimal covariance structure. A pharmaceutical example was also used to evaluate the performance of these fit statistics empirically. The LRT failed to decide which is a better model because LRT can only be used for nested models. SRMR, on the other hand, had congruent result as AIC and BIC and chose ARMA(1,1) as the optimal variance-covariance structure. In the second study, the author adopted a first-order autoregressive structure as the true within-subject V-C structure with variability in the intercept and slope (estimating [tau]00 and [tau]11 only) and investigated the consequence of misspecifying different levels/types of the V-C matrices simultaneously on the estimation and test of significance for the growth/fixed-effect and random-effect parameters, considering the size of the autoregressive parameter, magnitude of the fixed effect parameters, number of cases, and number of waves. The result of the simulation study showed that the commonly-used identity within-subject structure with unstructured between-subject matrix performed equally well as the true model in the evaluation of the criterion variables. On the other hand, other misspecified conditions, such as Under G & Over R conditions and Generally misspecified G & R conditions had biased standard error estimates for the fixed effect and lead to inflated Type I error rate or lowered statistical power. The two studies bridged the gap between the theory and practical application in the current literature. More research can be done to test the effectiveness of proposed SRMR in searching for the optimal V-C structure under different conditions and evaluate the impact of different types/levels of misspecification with various specifications of the within- and between- level V-C structures simultaneously.