Doctor of Philosophy, The Ohio State University, 2023, Educational Studies

As mobile learning (m-learning) becomes increasingly prevalent in education, it is recognized for its potential to enhance the overall quality of teaching and learning. Despite the many benefits, m-learning apps often experience low retention rates, which directly impede learners' benefit from using them and cause a waste of resources in app design, development, and maintenance. To investigate the critical factors influencing learners' acceptance of m-learning outside the classroom, this study introduced a novel model, the Mobile Learning Acceptance Determination (mLAD) Model, based on the Technology Acceptance Model and the updated DeLone and McLean Information System Success Model. Through the mLAD model, the study identified the critical app quality factors that influence learners' acceptance of m-learning. The moderating effects of the type of m-learning apps on learners' acceptance of m-learning were also revealed. An online questionnaire named the m-Learning Acceptance Questionnaire (mLAQ) was developed and disseminated through Amazon Mechanical Turk. A total of seven hundred forty-seven adult learners in the U.S. participated in the study. The descriptive statistical results of the examined factors revealed that m-learning apps available in the market demonstrate high mobility and content quality. Still, their interactivity and service quality could be improved. Furthermore, the results of the structural equation modeling analysis indicated that learners' two beliefs, perceived usefulness, and perceived ease of use, are the two essential determinants of learners' intention to use m-learning apps outside the classroom. Quality factors, such as content quality, interface design, mobility, and service quality, are the antecedents of learners' m-learning acceptance, given that they significantly and directly influence perceived usefulness and ease of use and indirectly impact learners' intention to use m-learning apps through learners' two beliefs. Through (open full item for complete abstract)

Committee: Ana-Paula Correia (Advisor); Minjung Kim (Committee Member); Richard J Voithofer (Committee Member) Subjects: Education; Educational Software; Educational Technology; Information Systems; Information Technology; Technology

Master of Arts (MA), Bowling Green State University, 2015, Psychology/Clinical

Objective. Social capital is generally defined as an individual's potential for tangible or social resources made available via interpersonal connections. Higher levels are related to a variety of positive health, well-being, and occupational outcomes. Social capital can be measured by a variety of indices in adulthood. Currently, the childhood precursors to adult social capital are relatively unknown. The current project tests a developmental-contextual model for both the measurement of social capital in adulthood and the childhood precursors that may impact its accumulation over the life course. Methods. 523 participants were surveyed at age 8 and again at age 48 as part of a 40-year prospective, longitudinal study. Participants completed measures of cognitive abilities, relationships with parents, peers, and spouses, information about personal traits, and beliefs and attitudes regarding social relationships. Structural Equation Modeling (SEM) was used to test the hypothesized model. Results. The proposed measurement model of social capital was generally supported, with latent domains of individual, interpersonal, socio-economic, and community level indices. Predictive models of childhood precursors for social capital differed by gender: prominent precursors were childhood aggression and cognitive ability for males, and childhood family religiosity for females. Conclusions. These findings suggest that a developmental-contextual model of social capital may account for the multiple indices of social capital and that its accumulation across the life course is different for males and females. Knowledge of precursors may be clinically helpful for fostering social capital growth in therapeutic settings.

Committee: Eric Dubow Ph.D. (Advisor); Carolyn Tompsett Ph.D. (Committee Member); Marie Tisak Ph.D. (Committee Member) Subjects: Clinical Psychology

Doctor of Philosophy, The Ohio State University, 2014, EDU Policy and Leadership

Under the framework of structural equation modeling (SEM), longitudinal data can be analyzed using latent growth models (LGM). An extension of the simple LGM is the multilevel latent growth model, which can be used to fit clustered data. The purpose of this study is to investigate the performance of five different missing data treatments (MDTs) for handling missingness due to longitudinal attrition in a multilevel LGM. The MDTs are: (1) listwise deletion (LD), (2) FIML, (3) EM imputation, (4) multiple imputation based on regression (MI-Reg), and (5) MI based on predictive mean matching (MI-PMM). A Monte Carlo simulation study was conducted to explore the research questions. First, population parameter values for the model were estimated from a nationally representative sample of elementary school students. Datasets were then simulated based on a two-level LGM, with different growth trajectories (constant, decelerating, accelerating), and at varying levels of sample size (200, 500, 2000,10000). After datasets are generated, a designated proportion of data points (5%, 10%, 20%) were deleted based on different mechanism of missingness (MAR, MNAR), and the five missing data treatments were applied. Finally, the parameter estimates produced by each missing data treatment were compared to the true population parameter values and to each other, according to the four evaluation criteria: parameter estimate bias, root mean square error, length of 95% confidence intervals (CI), and coverage rate of 95% CIs. Among the five MDTs studied, FIML is the only MDT that yields satisfactory bias level as well as coverage rate for all parameters across all sample sizes, attrition rates, and growth trajectories under MAR. It is also the only MDT that consistently outperforms the conventional MDT, LD, in every aspect, especially when missingness ratio increases. Under MNAR, however, estimates of the predictor effects on slopes become biased and coverage for those two paramet (open full item for complete abstract)

Committee: Richard Lomax (Advisor); Paul Gugiu (Committee Member); Eloise Kaizar (Committee Member) Subjects: Education; Statistics

PhD, University of Cincinnati, 2001, Medicine : Environmental Health Sciences

To achieve detailed understanding of an exposure-outcome association in public health studies, an investigator often needs to account for mediator(s). A mediator is a variable that occurs in a causal pathway from an independent to a dependent variable. The mediational model describes the associations among the exposure(s), mediator(s), and outcome(s). Statistics are needed to determine how much of the exposure-outcome association is due to a mediator. Although mediational models are widely applied in public health, sociological and psychological research, the statistical methods to define and test mediation effects are underdeveloped. The current available methods, path analysis and multi-step regression analyses, have some major limitations including: 1) lack of clear and meaningful definitions of mediation effects; 2) lack of significance testing procedures for the mediation effects; and 3) these methods have not been extended into a generalized form. The present study defined mediation effects, which allows for substantive accounting of the exposure-outcome process that is consistent across a class of generalized mediational models. The newly defined mediation effects have important epidemiological interpretations that are closely related to the concept of attributable risk (AR). Both linear and non-linear models were studied. Much attention has been given to the logistic mediational model due to its important role in the epidemiological studies. Asymptotic variance estimates for the mediation effects were derived using the multivariate delta method. Through Bayesian modeling and Monte Carol techniques, in particular, Markov chain Monte Carlo (MCMC), the posterior distributions for the mediation effects were estimated. Simulation studies, as well as case studies using a nationally representative database, compared the behavior of the asymptotic estimates and the non-informative Bayesian posterior estimates for the mediation effects of the linear and logistic medi (open full item for complete abstract)

Committee: Dr. Paul Succop (Advisor) Subjects: Health Sciences, General

Doctor of Philosophy, The Ohio State University, 2003, Psychology

This dissertation represents an investigation into the role of model complexity in structural equation modeling (SEM) and how traditional notions of model fit, which do not typically consider complexity, are inadequate for summarizing the success or failure of a model. Model fit is traditionally summarized in an index which communicates the agreement between a given model and a particular data set. However, demonstrating that a given model could have generated a given data set is not sufficient to show that it is a good model. Because complexity partially determines the ability of a model to fit data, model complexity should not be ignored when evaluating model fit. The importance of model complexity is examined in the SEM context by simulating correlation matrices and applying a series of models which differ in complexity. First, it is shown that models differing in the number of free parameters can fit the same random data differentially well by a simple criterion of good fit. Second, models with the same number of free parameters, but which differ in functional form, are found to fit the same data differentially well. Third, it is found that restrictions placed on parameter range have an impact on model complexity. Because traditional fit indices usually correct for model complexity due only to the number of free parameters, these findings have important implications for how models are assessed and compared in the SEM paradigm.

Committee: Robert MacCallum (Advisor) Subjects: Psychology, Psychometrics

PHD, Kent State University, 2021, College of Education, Health and Human Services / School of Lifespan Development and Educational Sciences

The purpose of this study was to investigate the relationship among self-efficacy, self advocacy, and multicultural counseling competency of school counselors currently practicing in the US. The research questions guided this study included (a) What are the direct and indirect influences of school counselor self-efficacy on multicultural counseling competence? (b) Is the relationship between self-efficacy and multicultural counseling competence mediated by self-advocacy for school counselors? The data were collected from 306 school counselors practicing in the US. Confirmatory Factor Analysis (CFA) and Structural Equation Model (SEM) were used in the data analysis in the study. The results suggested self-efficacy significantly predicted multicultural counseling competence among the US school counselors. The results of the hypothesized structural model also suggested that self-advocacy had a strong indirect effect on multicultural counseling competence mediated by self-efficacy. The results of the data analysis, discussions of the findings, implications of the current study, and limitations and future research directions are presented herein.

Committee: Jason McGlothlin DR (Committee Chair); Martin Jencius DR (Committee Member); Kelly Cichy DR (Committee Member) Subjects: Counseling Education; School Counseling

Doctor of Philosophy, The Ohio State University, 2020, Entomology

Urban areas are often considered adverse environments for wildlife communities given that the colonization and establishment of local species pools are disrupted by biotic and abiotic changes at different spatial scales such as the introduction of invasive species, periodic mowing, and changes in soil and air quality. Although the number of people residing in cities has increased in the last century, over 300 cities worldwide have shrunk due to prolonged economic decline and population loss, resulting in thousands of newly available greenspaces scattered throughout cities. Consequently, interest in urban greenspaces as sites for conservation has grown considerably, raising questions about the ability of these habitats to support wildlife. As novel ecosystems, urban areas represent a set of new challenges for local species pools, yet the mechanisms driving community assembly processes within cities is a major knowledge gap. My work focused on identifying species distributions, patterns, and processes leading to the successful establishment arthropods in cities. For this, I chose to work with beetle and spider assemblages as they are considered biological indicators of environmental changes at small and large spatial scales and are taxonomically and functionally diverse predatory groups. In Chapter 1, my objective was to determine how urban greenspaces management and design impacts Carabidae and Staphylinidae beetles using taxonomic and life-history trait approaches. I found that ecological and morphological traits were good indicators of how beetles were responding to greenspace management strategies. Most species were negatively associated with building structures, while undisturbed habitats supported more hygrophilous and brachypterous beetle populations. In Chapter 2, I investigated the importance of local and landscape characteristics on spider communities using taxonomic and functional diversity approaches. I found that Pardosa milvina (Lycosidae) was the mo (open full item for complete abstract)

Committee: Mary Gardiner PhD (Advisor); Luis Cañas PhD (Committee Member); Andrew Michel PhD (Committee Member); Robert Gates PhD (Committee Member); William Symondson PhD (Committee Member) Subjects: Agriculture; Animals; Behavioral Sciences; Biology; Biostatistics; Ecology; Entomology; Environmental Science; Forestry; Molecular Biology; Soil Sciences; Urban Forestry; Urban Planning; Wildlife Conservation; Wildlife Management

Doctor of Philosophy (PhD), Ohio University, 2020, Educational Research and Evaluation (Education)

Two main research problems were addressed in the current study. First, the researcher explored the impact of e-mail prenotification, follow-up reminders and of mixed-mode design on survey response rates in Saudi Arabia among four conditions when applying a 2 (pre-notification: Yes, No) × 2 (Follow-up: E-mail, WhatsApp) between-subjects factorial design. Further, this study investigated the impact of including the phrase, “all I need is 10 more people,” during survey distribution. Results indicated that using both e-mail prenotification and follow-up reminders simultaneously, as well as multiple follow-up reminders in the form of both email and social media applications increased response rate. Further, using the phrase “all I need is 10 more people” during the second-follow reminder both elevated the response rate and provided support from the university to the researcher. Second, the researcher demonstrated whether the analytic choice between MR and SEM affected the results when examining the factors impacting the research productivity of faculty members in Saudi Arabia. Results indicated that using either MR or SEM delivered different results in terms of significant predictors and the model's overall explained variance. Further, differential outcomes produced by the various SEM models employed illustrate how the incorrection specification of formative (causal) indicators can result in worse data-fitting models. Implications for selecting analytic procedures are discussed.

Committee: Gordon Brooks (Committee Chair); Yuchun Zhou (Committee Member); Charles Lowery (Committee Member); Lijing Yang (Committee Member); Anirudh Ruhil (Committee Member) Subjects: Education

Master of Science, The Ohio State University, 2017, Human Ecology: Human Development and Family Science

Social-emotional competence is critical to young children's success in social and academic settings across the lifespan. Non-parental primary caregivers are important socializers of children's social emotional development, particularly through the ways they respond to children's negative emotions. Despite this, little research has examined predictors of responsiveness or the ways they interact to influence responsiveness in samples of non-parental caregivers. The detrimental influence of elevated depression and stress on individual's affect and interactions has been consistently documented in research; additionally, previous research suggests that work-family conflict may decrease responsiveness in parent samples. This study examined how depression and stress was associated with family child care providers' responsiveness, and the influence of family functioning as a mediator. Direct and indirect associations were examined utilizing structural equation modeling with a national survey of 888 small licensed family child care providers from across the United States. This study found when family child care providers perceived higher levels of general stress, they reported that they utilized less positively-focused reactions, expressive encouragement, and positive social guidance; they did not report using more negative reactions or negative social guidance. When family child care providers reported higher levels of general stress and depression, they reported lower levels of family functioning. In turn, family functioning was significantly associated with each responsiveness measure, except for negative social guidance. Providers who reported higher levels of family functioning also reported responding to children in more positive ways. Additionally, higher levels of family functioning were associated with less negative reactions from family child care providers. Bootstrap analysis results found some mediation effects from family functioning for stress and responsivenes (open full item for complete abstract)

Committee: Cynthia Buettner PhD (Advisor); Suzanne Bartle Haring PhD (Committee Member) Subjects: Early Childhood Education; Education Policy; Families and Family Life; Gender Studies; Individual and Family Studies; Mental Health; Preschool Education; Teacher Education; Womens Studies

Doctor of Philosophy, The Ohio State University, 2017, Social Work

Introduction: The excessive consumption of alcohol by college students is a major public health problem in the U.S. Heavy alcohol use has been linked to numerous consequences ranging from less serious effects (i.e.., hangovers) to death. Decades of research have linked certain beliefs, attitudes or motivations to drinking behavior but intensive prevention efforts based on these ideas have done little to mitigate this issue. Much of the past research has focused on the interplay of cognitive factors (i.e., expectancies, motivations) and typical drinking patterns (i.e., quantity or frequency of drinking during the past year). Unfortunately, examining the relationship between “general” motives, expectancies, or use of protective strategies and “typical” drinking is not adequate to understand behavior as it occurs. Therefore, the need to understand drinking at the event-level is critical. To this end, this study examined a conceptual model of college students' drinking events in order to determine the potential mediating effect of drinking motives and protective behavioral strategies in the relationship between alcohol expectancies and event-level alcohol use and consequences. Methods: An existing dataset containing information about 2,279 college student drinking events was analyzed for this study. Students completed surveys during the administration of a commercial online alcohol course during 2010 and 2011. These surveys contained measures of typical alcohol expectancies, drinking motives, and use of protective behavioral strategies. Students also provided detailed information about their last drinking event that occurred within seven days prior to the course. A theoretical model that examined the mediating influence of these cognitive factors and typical use of protective strategies on event-level alcohol use was analyzed with structural equation modeling. Results: The hypothesized causal ordering was supported by the findings. Both typical use of protective strate (open full item for complete abstract)

Committee: John Clapp Ph.D (Advisor); Thomas Gregoire Ph.D (Committee Member); Alicia Bunger Ph.D (Committee Member) Subjects: Behaviorial Sciences; Higher Education; Public Health; Social Work

Doctor of Philosophy, The Ohio State University, 2015, Psychology

Categorical confirmatory factor analysis (CCFA) and item response theory (IRT) are two closely related methods of analyzing categorical measurement data. Though their historical development and current usage are often different, they are mathematically related and differ primarily in the details of implementation. One historical difference that is beginning to fade is the standard approaches to assessing model fit in each framework, especially with regard to omnibus model fit. The root mean square error of approximation (RMSEA) is a popular omnibus fit statistic in structural equation modeling (SEM) generally and limited-information CCFA as a special case. Recent developments have resulted in the creation of a similar statistic for use in IRT, a full-information counterpart to limited-information CCFA. Much work has been devoted to determining acceptable standards of model fit, often resulting in recommended cutoffs, though recent research suggests that universal application of such rules of thumb is inappropriate. Additionally, these studies have largely investigated the RMSEA in the context of continuous variables.The present study compares the estimated RMSEAs for limited- and full-information solutions for categorical data in light of recent research and provides suggestions for approaching fit in psychometric modeling.

Committee: Michael Edwards Ph.D (Advisor); Paulus De Boeck Ph.D (Committee Member); Min-Jeong Jeon Ph.D (Committee Member); Duane Wegener Ph.D (Committee Member) Subjects: Psychology; Quantitative Psychology; Statistics

PhD, University of Cincinnati, 2007, Medicine : Epidemiology (Environmental Health)

Pediatric studies of associations among obesity, asthma and allergic sensitization are inconclusive. Body mass index (BMI) is commonly used to classify study participants as obese. BMI, however, does not reflect central fat distribution, a known risk factor for metabolic disorders. This research tested the hypothesis that central fat distribution increases risk for asthma and allergic sensitization, measured using the skin prick test for aeroallergens. Measures of central fat distribution included waist circumference, waist to height ratio (WHtR) and conicity index. Childhood studies rarely utilize these alternative measures for characterizing the relationship between obesity and asthma or allergic sensitization. In this study, alternative measures of obesity were compared with BMI percentiles; the association between obesity and asthma in children with allergic rhinitis was investigated; the association between obesity and allergic sensitization in children with and without asthma, rhinitis and eczema was explored. To account for the effects of several obesity measures on the above outcomes, a structural equation model was developed. Results show that BMI percentiles are discordant with 18-46% of children classified as obese using the alternative measures. A high WHtR or conicity index increases likelihood for asthma in children with allergic rhinitis. Prevalence of a high WHtR consistently increases from allergic rhinitis to mild asthma to moderate/severe asthma. Sensitization to aeroallergens appears to protect against an asthma diagnosis (OR=0.6; 95% CI=0.4-0.9), but is positively associated with rhinitis (OR=8.1; 95% CI=4.9-13.5) and eczema (OR=1.9; 95% CI=1.0-3.4). In children with rhinitis, those with a moderate WHtR are twice as likely to have increasing positive skin prick tests (OR=1.9; 95% CI=1.1-3.3) compared to low WHtR. In children with asthma, obesity appears to protect against increasing positive skin prick tests per BMI percentiles (OR=0.5; 95 (open full item for complete abstract)

Committee: Dr. Kim Dietrich (Advisor) Subjects:

Doctor of Philosophy, The Ohio State University, 2009, Rural Sociology

This study examines the role of a community or place in the social control of crime. The objectives of this study are to answer questions of the where, the why and the how of crime rate variations in rural communities, and if these are associated with different degrees or levels of social control. For the purpose of this dissertation, a community or place is defined as a locality where people interact with each other and share, at least to some extent, a common identity. This study uses a macro-level perspective to study rural community and crime. The unit of analysis is the county, which serves as a proxy for community. It is assumed that county-level social structural and socioeconomic characteristics determine its social integration and social control. Following the tradition of social disorganization theory originating from Shaw and McKay, five macro-level social structural and socioeconomic status characteristics are employed and tested. Social disorganization theory argues that crime is associated with social structural and socioeconomic characteristics that negate or reduce the ability of local groups and individuals to control criminal behavior. This study focuses both on the spatial and temporal differences of nonmetropolitan counties and its consequence for variations in crime rates. A sample of 1,541 nonmetropolitian counties is used in this study. Data of county crime rates is obtained from the Uniform Crime Report (UCR) of FBI. Data of county social structural and socioeconomic characteristics is obtained from the 1990 and 2000 Census of the Population. The Rural-Urban Continuum Codes come from the Economic Research Service (ERS) of the United States of Department of Agriculture, which provides information for the classification of metropolitan and nonmetropolitan counties. This study adopts multiple regression and structural equation model analyses to test various hypotheses. There are several findings: (1) We confirmed that social disorganization pers (open full item for complete abstract)

Committee: Joseph F. Donnermeyer (Advisor); Linda Lobao (Committee Member); Flinn William (Committee Member) Subjects: Social Structure

Doctor of Philosophy, The Ohio State University, 2003, Social Work

Juvenile offending has been a nationally recognized persistent social problem. Current interventions are still distant from comprehensive and holistic resolutions in preventing and decreasing delinquent behaviors. One critical limitation of the existing intervention strategies for decreasing juvenile delinquency is the exclusion of religious factors. The main purpose of this dissertation is to reexamine the relationship between religiosity and delinquency in a multivariate context by adopting advanced research methods and extending earlier work with a sample from ADD Health data. Additionally, in recognition of the importance of temporal order between exogenous variables and endogenous variables, exogenous variables are chosen from the wave 1 data and endogenous variables were taken from the wave 2 data for the present study. The use of two wave data greatly reduces the threat of spuriousness. As the theoretical framework of the present study, the social control theory is elaborated upon by incorporating the basic concept of delinquent peer influence from the differential association theory. Utilizing the structural equation model (SEM), the theoretical model examines both the direct effects of adolescent religious belief and its indirect effect through other intervening factors from social control theory and differential association theory. Also, based on previous studies that the effects of religiosity vary across different types of delinquent behavior, the study examines the effect of religiosity separately for minor offenses and serious offenses. The results indicate that adolescent religiosity has a statistically significant direct and independent effect on serious delinquent behaviors. Yet the results are somewhat different for minor offenses. The persistent and significant effect of adolescent religiosity on family, peers, and schools provides support that the effect of religious belief on delinquency can be extensive in deterring delinquent behaviors through (open full item for complete abstract)

Committee: Denise Bronson (Advisor) Subjects: Social Work

Doctor of Philosophy (PhD), Ohio University, 2011, Curriculum and Instruction (Education)

Athletic coaches have the potential to be among the most influential people in a young person's life and athletes often idolize their coaches. The impact a coach has on an athlete endures psychologically, physically, and emotionally. This study embraced the Sport Commitment Model (Scanlan, T. K., Carpenter, P. J., Schmidt, G. W., Simons, J. P., & Keeler, B., 1993a; Scanlan, T. K., Russell, D. G., Magyar, T. M., & Scanlan, L. A.,, 2009) as a theoretical framework to understand the antecedents of coaches' commitment to coaching. The goal of this study was to examine the viability of the Coaches' Commitment Model (CCM) with soccer coaches. The theoretical framework of the SCM provided a mechanism to understand the determinants of soccer coaches' commitment to coaching, as the modified measurement model met satisfactory model fit (χ2 = 753.5 [df = 215], CFI = .954, NNFI = .946, RMSEA = .039, and SRMR = .0388). While the structural model failed to converge, this does not mean the SCM is not a viable theoretical framework for coaches. The theory behind the models (i.e., SCM and CCM) is that enjoyment, involvement alternatives/other priorities, personal investments, social constraints, involvement opportunities/valuable opportunities and social support predict coaches' commitment. The soccer coaches in the study seem to express their lifelong involvement in sports, thus their commitment to coaching. Overall, they have been long time athletic participants and consider themselves more than just novice coaches. Additionally, the opportunity to coach and work with athletes was the strongest predictor of coaches' commitment. Furthermore, these coaches not only value the opportunities to work with their athletes, they enjoy coaching.

Committee: Dianne Gut (Committee Chair); George Johanson (Committee Member); David Carr (Committee Member); Ronald Quinn (Committee Member) Subjects:

Doctor of Philosophy, Case Western Reserve University, 2011, Social Welfare

The purpose of this dissertation was two fold: 1) to examine the relationships between depressive symptoms and alcohol use among middle-aged and early older adults and 2) to examine whether retirement status moderates the differences in the relationships. For examining these aims, this study utilized a multi-group cross-lagged structural equation modeling using a pane data of the Wisconsin longitudinal study (N=3,204). No support for a mutually reinforcing relation between depressive symptoms and alcohol consumption was found in the current study. Higher levels of depressive symptoms did not lead to higher alcohol consumption among early older adults, and alcohol consumption was not associated with later depressive symptoms by analyzing the cross-lagged SEM. Therefore, the nature of the causal processes contributing to the obtained pattern of relationships between depressive symptoms and alcohol consumption cannot be determined by this study. These results may reflect that the mechanisms responsible for co-morbidity of depressive symptoms and alcohol-related problems could vary across individuals and represent etiologically distinct subgroups. There was no significant moderating effect of retirement status in the relationship between depressive symptoms and alcohol use by analyzing the multi-group SEM. Several caveats should be taken into consideration when interpreting the findings of this study. Although this study is a population-based study, the sample was limited to a certain population who graduated from Wisconsin high schools in 1957, and thus, the results have a limitation in terms of generalization. Despite several limitations, the study makes contributions in several areas. Given the fact that a longitudinal design is the optimal research endeavor when examining the relationships between depressive symptoms and alcohol use, this study will provide invaluable information for understanding the relationships. This study has also provided a rigorous test of (open full item for complete abstract)

Committee: Kathleen Farkas PhD (Committee Chair); Kathryn Adams PhD (Committee Member); Diana Morris PhD (Committee Member); Meeyoung Min PhD (Committee Member) Subjects: Social Work

Doctor of Philosophy, Case Western Reserve University, 2009, Epidemiology and Biostatistics

Most diseases of interest to modern genetic epidemiologists are complex both in their etiology and measurement. That is, they result from a complicated interplay of various environmental and genetic factors, and they are subject to fuzzy, noisy and often multidimensional disease definitions. Although complex diseases are inherently multivariate, it is often difficult to see how multivariate methods may be used in family data. For example, there are several contradictory claims in the literature about the asymptotic distribution of the multivariate variance component likelihood ratio test for linkage analysis. We show that the previous claims are not correct, but computational efficient algorithms may be used to find the distribution. However, the likelihood ratio test is not robust to non-normality in this context, so several robust score tests for multivariate linkage analysis are developed. Via extensive simulations, we explore the statistical properties of these tests. Finally, a framework for using structural equation models (SEM) in family data is developed. This framework includes both a latent measurement model and a structural model with covariates. This allows for a wide variety of models, including latent growth curve models. It is shown how variance components such as polygenic, environmental and genetic variance components can be included in the SEM.

Committee: Catherine Stein PhD (Advisor); Robert Elston PhD (Committee Chair); Xiaofent Zhu PhD (Committee Member); Ralph O’Brien PhD (Committee Member) Subjects: