The goal of this research is to construct causal gene networks for genetical genomics experiments using expression Quantitative Trait Loci (eQTL) mapping and Structural Equation Modeling (SEM). Unlike Bayesian Networks, this approach is able to construct cyclic networks, while cyclic relationships are expected to be common in gene networks. Reconstruction of gene networks provides important knowledge about the molecular basis of complex human diseases and generally about living systems. In genetical genomics, a segregating population is expression profiled and DNA marker genotyped. An Encompassing Directed Network (EDN) of causal regulatory relationships among genes can be constructed with eQTL mapping and selection of candidate causal regulators. Several eQTL mapping approaches and local structural models were evaluated in their ability to construct an EDN. The edges in an EDN correspond to either direct or indirect causal relationships, and the EDN is likely to contain cycles or feedback loops. We implemented SEM with genetics algorithms to produce sub-models of the EDN containing fewer edges and being well supported by the data. The EDN construction and sparsification methods were tested on a yeast genetical genomics data set, as well as the simulated data. For the simulated networks, the SEM approach has an average detection power of around ninety percent, and an average false discovery rate of around ten percent. / Ph. D.

In this study, 236 heavy-drinking college students completed measures of self-efficacy for limiting drinking, specific coping skills for limiting drinking, outcome expectancies associated both with expected effects of drinking and expected effects of limiting drinking, and retrospective drinking behavior. Confirmatory factor analyses were conducted to examine reliability and validity, as well as serving as a pre-requisite for structural equation modeling (SEM). Results were generally consistent with predictions and supported the distinction between self-efficacy and outcome expectancies. One notable exception was that positive expectancies for limiting drinking did not load heavily on the predicted expectancy construct. Three models predicting drinking were compared utilizing SEM. The first was a model in which all constructs predicted drinking with no indirect effects. The second was based upon the ideas of Kirsch (1995) and predicted that outcome expectancies influence self-efficacy judgments. The third was based upon Bandura's (1986) theory and predicted that self-efficacy judgments would instead influence outcome expectancies. Both the models based on Kirsch and Bandura appeared to better fit the data than the model with no indirect effects. Differences in model-fit between models based on Kirsch and Bandura were not large, but slightly supported the Kirsch model. Additional analyses also supported the importance of outcome expectancies in predicting drinking behavior. Implications for theory and future directions for research are discussed. / Ph. D.

High rates of medical errors are well documented within the healthcare industry. Nurses, in particular, play a critical role in the quality and safety of healthcare services. Fatigue is a factor that has been linked to stress, safety, and performance decrements in numerous work environments. Within healthcare, however, a comprehensive definition of fatigue encompassing multiple dimensions has not been considered, but is warranted since nurses perform tasks consisting of diverse physical and mental activities. As such, "total fatigue" was examined, as were interactions between its underlying dimensions and the effects of these dimensions on performance, in the context of actual and simulated nursing work. In a survey study (Chapter 2), registered nurses reported relatively high levels of mental, physical, and total fatigue, and higher levels of fatigue were associated with perceived decreases in performance. Work environment variables, such as work schedule or shift length, were also related to differences in reported fatigue levels. An experimental study investigated causal effects of mental and physical fatigue on mental and physical performance (Chapter 3). Mental fatigue affected a measure of mental performance, and physical fatigue had a negative effect on measures of physical and mental performance. A multidimensional view of fatigue that considers direct and crossover effects between mental and physical dimensions of fatigue and performance is relevant when quantifying effects of fatigue on performance. A model of the relationships between fatigue dimensions and performance in nursing was developed using structural equation modeling techniques (Chapter 4). The model supported the existence of a total fatigue construct that is directly affected by mental and physical fatigue levels. The final model also provides quantitative path coefficients defining the strength of relationships between mental and physical dimensions of fatigue, total fatigue, and mental and physical performance. The current research provides an increased understanding of fatigue levels in registered nurses across work environments, as well as the underlying causal mechanisms between dimensions of fatigue and performance decrements. The findings and the final model can aid in designing interventions to reduce or eliminate the contributions of fatigue to the occurrence of medical errors. / Ph. D.

This study seeks to offer a meaningful statement about the relative importance of self-processes and activated social capital in the process that links high school students and educational outcomes. The conceptual model for the study draws on the large and diverse body of research that aims to understand the process and effects of the interaction of the person and his or her environment. It is hypothesized that adaptive self-processes and social capital are positively related to school engagement, educational aspirations, and actual performance in school; and, that these factors mediate the effects of family background and other potential social capital on educational outcomes. The data for this study were obtained from a sample of N=1,176 in grades 9-12 from three school divisions in Virginia. Students completed the School Relationships and Experiences Survey (SRES), an instrument designed for this study. The study uses structural equation modeling (SEM) to model the relationships between the variables of interest. Data were analyzed using LISREL 8.3 (Jöreskog & Sörbom, 1993). The covariance structure models tested include both single-indicator and multiple-indicator constructs. The analysis follows the two-step procedure suggested by Anderson and Gerbing (1988). First, a measurement model was tested using confirmatory factor analysis to develop a model with acceptable fit to the data. In step two, the theoretical model of interest was specified as an a priori model. This theoretical model was then tested and revised until a theoretically meaningful and statistically acceptable model was found. In conclusion, the results of the analyses are discussed, and possible explanations for the results are proposed. Directions for future research are outlined, including the need for cross-validation of this model on additional samples of high school students. Social capital has previously been conceptualized primarily as family resources and parent-child relationships. This study provides promising initial evidence that activated social capital (education-related support received by adolescents from the broader social network) has stronger and more meaningful effects on academic engagement and other educational outcomes than the more passive indicators of social capital used in previous research. This work was supported by a doctoral fellowship from the Social Sciences and Humanities Research Council of Canada (SSHRC) and a grant from the ASPIRES program at Virginia Tech. / Ph. D.

The current study drew on a large and diverse body of literature on social capital and aimed to understand its role in the process of transition from college to work. In particular, this research studied the cumulative effects of social capital formed in high school years and college years and examined its relationship with job outcomes. The study used the National Education Longitudinal Study (NELS) to examine whether early investment in the social capital of young adolescents produced better job outcomes in their adulthood. Families and schools were two primary sources of social capital considered in the current study. Parental involvement in a young person's life, extra-curricular activities and participation in volunteer organizations were some of the forms of social capital hypothesized to influence job outcomes after college. Structural equations modeling was used to trace the effects of the presence of social capital as early as the 8th grade in shaping student's later career status. The longitudinal data measured social capital beginning in the 8th grade and every 2 years thereafter, so that the cumulative effects of the social capital resources were investigated. Overall, the hypothesized model was found to fit the data and the findings have suggested a set of positive and direct effects of social capital on job outcomes. / Ph. D.

In this study, conceptual models of young adult citizenship participation eight years post high school using Putnam's (1993) theory of social capital are tested. This study uses a longitudinal database. Specifically, the models posit that accumulation of more social capital resources during adolescence will be associated with more positive future life expectations related to educational, occupational, and family life outcomes. Subsequently, these more positive future life expectations and social capital resources will be associated with a higher degree of citizenship participation (as measured by voting behavior, community integration and volunteerism) as young adults. Measures from the 1990, 1992 and 2000 waves of the National Education Longitudinal Study (NELS:88) dataset were used. The models were examined using structural equation modeling procedures. Findings indicate that school and community social capital had mixed results regarding their effects on the final outcome variables. Teacher interactions were not found to have a statistically significant direct effect on any of the three citizenship participation outcome variables. But this result should be viewed with caution due to possible measurement problems with the construct of teacher interactions. Peer interactions did have a statistically significant, direct, effect on community integration and volunteerism, but it did not have a direct effect on voting behavior. Participation in extracurricular school activities did have a small, direct, effect on all three outcome variables. Participation in community sports activities had a positive direct effect on community integration, a small negative effect on voting behavior, and no statistically significant effect on volunteerism. Finally, participation in other types of community activities had the largest direct effects on all three citizenship participation outcome variables. The adolescent social capital variables of teacher interactions, peer interactions, and participation in other types of community activities for tenth graders did have positive, direct effects on future life expectations for twelfth graders. Extracurricular activities participation and community sports participation were not found to have an effect on future life expectations. Future life expectations were found to have a very small, but statistically significant, positive, direct effect on all three citizenship participation variables. These findings indicate that some adolescent social capital resources can have positive direct effects on the long term, young adult, educational outcome of citizenship participation. But it varies by type of social capital resource and by the particular aspect of citizenship behavior being measured. / Ph. D.

Although anxiety disorders and conduct problems often co-occur in children and adolescents, literature describing the effects of such co-occurrence is mixed. There is evidence that symptoms of anxiety disorders may mitigate symptoms of conduct problems (buffering hypothesis) or may exacerbate symptoms of conduct problems (multiple problem hypothesis). A dual-pathway model has been proposed that suggests several possible etiological or risk processes that may differentiate these pathways (i.e., the buffering hypothesis or the multiple problem hypothesis) (Drabick, Ollendick, & Bubier, 2010). Executive functioning is one factor that has been identified that may differentially confer risk to the proposed pathways; however, little research has been done investigating its role. The purpose of the present study was to evaluate the dual-pathway model by determining whether executive functioning abilities contribute to differentiating those youth for whom anxiety exacerbates conduct problems from those for whom anxiety mitigates conduct problems. Specifically, the study sought to examine if executive functioning moderated the effect of anxiety symptom severity on conduct problems. Latent variable structural equation modeling (SEM) was used to analyze the data of 221 youth aged 9 to 16 in a residential treatment center who completed a full neuropsychological evaluation. Results of the study failed to support the hypothesis that executive functioning moderates the effect of anxiety on conduct problems. Furthermore, a structural equation model without an interaction between executive functioning and anxiety was found to fit the data better than a model with an interaction between those variables. Overall, the study found that executive functioning abilities could not distinguish youth for whom anxiety exacerbates conduct problems from youth for whom anxiety mitigates conduct problems. Recommendations for future research in light of the limitations of the current study, as well as remaining gaps in the literature, are discussed. / text

Much of the research on childhood sexual abuse (CSA) has utilized demographic factors (e.g., the gender, age, and race/ethnicity of the child victim) and specific characteristics of the abusive experience (e.g., the severity of abuse, the relationship between the perpetrator and the child victim, the child's age at the onset of abuse, etc.) as predictors of children's adjustment. Unfortunately, the ability of such factors to predict children's adjustment consistently across studies has been limited. Alternatively, while it has been recognized that survivors of CSA often hail from dysfunctional family environments and that the disclosure of CSA frequently results in emotional distress for caregivers, less research has focused on how these factors may serve to influence sexually abused children's adjustment. This study included 163 sexually abused youth and their non-offending caregivers who completed measures of psychological adjustment and the family environment. Demographic information and abuse-related characteristics were also collected. Children's age, gender, and race/ethnicity were not significantly predictive of emotional adjustment. Structural equation modeling was used to examine the direct and indirect effects of total abuse severity, caregiver distress, and children's perceptions of their family environments as predictors of emotional adjustment. Results indicated that both children's perceptions of their family environments and total abuse severity were significantly predictive of children's emotional adjustment, whereas caregiver distress did not have significant direct or indirect effect on adjustment. Finally, the implications of these findings for future research and the development of effective interventions for sexually abused children and their families was discussed

Numerous studies have indicated racial and ethnic disparities in the vocational rehabilitation (VR) system, including differences in eligibility, services provided, and employment outcomes. Few of these studies, however, have utilized advanced multivariate techniques or latent constructs to measure quality of employment outcomes (QEO) or tested hypothesized models for the relationship between race, personal history characteristics, and VR outcomes. Furthermore, few VR disparities studies have examined southwestern states such as Texas, which has large Hispanic and Black populations. The purpose of this study was to utilize structural equation modeling (SEM) to examine several implied conceptual models for the relationship between race, personal history characteristics, and VR outcomes for White, Black, and Hispanic participants in the Texas VR system. The implied conceptual models were tested for goodness of fit and multiple-group invariance. A measurement model for QEO, a latent construct, was tested and used in the study. QEO was measured by three indicator variables and evaluated using confirmatory factor analysis. A MIMIC model was tested to assess racial/ethnic variation in QEO. The MIMIC results were compared to a multiple regression approach. In addition, a path model and logistic regressions were conducted to assess racial variation in VR closure status among consumers who were unemployed at application to VR. All models were retested with an independent sample to assess predictive validity. The study results indicated good model fit and measurement invariance for the QEO construct. The structural model for race, personal history characteristics, and QEO indicated moderate model fit. It also indicated interaction effects for race by gender and for race by public support. The MIMIC model results suggest that QEO decreased for Blacks and Hispanics compared to Whites. Furthermore, the MIMIC results, which utilized QEO as an endogenous variable, differed from the multiple regression findings, which utilized one criterion. The multiple regression findings indicated no statistically significant difference between Blacks and Whites. The path model for race and VR closure status indicated poor model fit. The logistic regression indicated no racial/ethnic differences in VR closure status. Several model estimates did not cross-validate. Study limitations and suggestions for future research are described. / text

Thus far, few research studies have examined the practice of integrated speaking test tasks in the field of second/foreign language oral assessment. This dissertation utilized structural equation modeling (SEM) and qualitative techniques to explore the relationships among topical knowledge, anxiety, and integrated speaking test performance and to compare the influence of topical knowledge and anxiety, respectively, on independent speaking test performance and integrated speaking test performance. Three instruments were employed in this study. First, three integrated tasks were derived from TOEFL-iBT preparation materials, and three independent tasks were developed specifically for this research study. Second, four topical knowledge tests (TKTs) were constructed by six content experts and validated on a group of 421 Taiwanese EFL learners. Third, the state anxiety inventory (SAI) from the State-Trait Anxiety Inventory was adopted. A total of 352 Taiwanese EFL students were recruited for the official study. At the first stage, they filled out the personal information sheet and responded to the TKTs. At the second stage, they took two independent tasks for which they spoke without input support, responded to an SAI, performed two integrated tasks in which they orally summarized the textual and auditory input given to them, and completed another SAI. Finally, 23 volunteers took part in follow-up interviews. The quantitative data were analyzed using the two-step SEM approach and the interview data were examined using a series of qualitative techniques, leading to five primary findings. First, topical knowledge and anxiety both strongly influenced the integrated speaking performance, though in an opposite manner. Second, topical knowledge did not significantly affect anxiety. Third, the effect of topical knowledge on independent speaking performance and integrated speaking performance varied depending on the topics of the tasks. Fourth, the impact of anxiety on independent speaking performance and integrated speaking performance also differed according to the topics of the tasks. Fifth, participants were overwhelmingly positive about the integrated tasks. In light of the findings, several implications are proposed for second/foreign language oral assessment theory, research methodology, and practice. / text