This study explored the emerging Self-Imposed Activity Limitation (SIAL) theory among community dwelling elders. This theory was examined using the proposed Aging Well Conceptual model which was guided by Baltes' Selection, Optimization with Compensation model, Markus and Nurius' Envisioned Possible Selves theory, Kuypers and Bengtson's Social Breakdown Syndrome model, Bandura's Self-Efficacy theory, and Rotter's Locus of Control theory. The objectives of this study were to explore the relationships among multiple variables in a hierarchical model and to examine the explanatory power of the SIAL variables in predicting elders' well-being.A correlational descriptive design with a causal modeling approach was used employing Structural Equation Modeling (SEM) techniques. The Aging Well model was tested through a secondary analysis of the National Survey of Midlife Development in the United States (MIDUS) database selecting respondents aged 65-74 years.Two research questions guided this study. Research question one, how well does the Aging Well model fit with empirical sample data, was explored. The Aging Well model statistically approximated the MIDUS data after theoretical and statistical modifications and explained 76% of the variance of elder's well-being. The mediating effects of SIAL variables were determined by nested alternative model testing. Research question two, are the proposition statements in the Aging Well model valid, and was demonstrated empirically by the expected patterns of correlation and covariance among most of the variables in the Aging Well model.SIAL as a composite factor had a large positive effect on elder's well-being. Elders' perceived constraints and perception of aging had no direct effect on well-being. The influences of these two factors on well-being were mediated by a common factor, SIAL. These findings supported the emerging SIAL theory by suggesting that the optimal use of SIAL would lead to adaptive outcomes promoting elders' well-being. In addition, SIAL mediated the effects of elders' sense of control and perception of aging on well-being. The full range of SIAL could not be examined due to limitations inherent in secondary data analysis.

Intercollegiate athletics are a prominent feature of American higher education. They have been characterized as the “front door” to the university due to their unique ability to draw alumni and other supporters to campus. It is often supposed that the exposure from high-profile athletics produces a number of indirect benefits including greater institutional prestige. Such exposure comes at a cost, however, as most Division I athletics programs are not financially self-sufficient and receive institutional subsidies to balance their budgets. At present, it is unclear how institutions budget for athletics subsidies or whether the recent increases in subsidies have impacted the overall financial picture of colleges and universities. Prior research has shown that athletics subsidies and student tuition and fees are not significantly correlated for public Division I institutions, which suggests the possibility that institutions have reallocated funds from other core areas to athletics. In this dissertation, the relationship between athletics subsidies and one of the most important core areas of the university – education and related activities – was examined. This relationship was investigated using fixed-effects structural equation models to analyze a panel dataset of public Division I institutions. It was found that total athletics subsidies (school funds and student fees) per student and education and related spending per student were positively correlated. This suggests that rather than decrease educational spending, institutions that increase total athletics subsidies have simultaneously increased their educational expenditures. However, in the analyses involving the more restrictive definition of athletics subsidies, it was shown that athletics subsidies from school funds was not correlated with educational spending. The results also provided some evidence that differences in the relationship between athletics subsidies and educational spending exist according to Carnegie classification and level of athletics competition. The findings from this study have a number of implications for higher education policy and future research. The absence of a negative relationship between athletics subsidies and educational spending suggests that athletics subsidies are not associated with decreases in educational spending that could ultimately harm the quality of education provided by colleges and universities. Furthermore, the existence of a positive correlation between athletics subsidies and educational spending and the fact that core revenues were controlled for in the models suggest the possibility that institutions have redirected funds from other areas to support education and athletics.

Although research is unequivocal concerning the important role of parenting in the prediction of a range of youth psychosocial outcomes, few empirical studies have examined potential contributions of parental individual differences factors to variability in parenting behaviors. Among the few studies that have, individual differences in affective dimensions of temperament (i.e., Negative Temperament [NT] and Positive Temperament [PT]) and executive functioning (EF) have individually emerged as potential key processes underlying parenting behaviors; however, they have yet to be examined jointly. Thus, using a latent variable approach, within a racially and ethnically diverse community sample of 166 parents, the current study examined the joint and interactive contribution of temperament and EF in the explanation of parenting. Further, despite conceptual overlap, parenting research has historically employed two distinct conceptual approaches: parenting practices and styles. The current study thus fitted a single integrative three-factor model (i.e., positive parenting, negative parenting, and corporal punishment) of parenting behaviors that included both styles and practices. Results suggested that parenting behaviors can be conceptualized within a single, three-factor model, allowing for the incorporation of historically distinct conceptions of parenting. Further, results revealed that affective dimensions of temperament and EF were uniquely but differentially associated with all parenting domains and EF moderated the associations between both NT and PT and positive parenting. All told, the current study provides support for an integrative model of parenting behaviors and parental temperament and EF, and their interaction, as potential critical processes associated with individual variability across parenting behaviors.

Previous literature provides an overview of the multiple relationships between alcohol use, protective behavioral strategies (PBS), alcohol-related negative consequences, depression, and sleep problems among college students, as well as differences by individual level characteristics, such as age, gender, and race/ethnicity. Several studies have found that specific demographic groups of students are more likely to reach a higher blood alcohol content (BAC) when “partying” or socializing (Turner, Bauerle, & Shu, 2004; Crotty, 2011). A variety of studies have also confirmed the positive relationship between high blood alcohol content and experiencing alcohol-related negative consequences (Turner, et al., 2004; Martens, Taylor, Damann, Page, Mowry, & Cimini, 2004; Borden, Martens, McBride, Sheline, Bloch, & Dude, 2011; Crotty, 2011). Additional studies have explored the role that protective behaviors play in the alcohol consumption-negative consequences relationship (Martens et al., 2004; Borden et al., 2011; Haines, Barker, & Rice, 2006; Martens, Martin, Littlefield, Murphy, & Cimini, 2011). These studies conclude that the frequency of protective behavior use and the number of these behaviors that are used when consuming alcohol are associated with the likelihood of a student experiencing negative consequences. Specifically, the negative relationship between protective behavior use and likelihood of experiencing negative consequences as a result of binge drinking is stronger for students who rarely use protective behaviors (Martens et al., 2004). Recent studies have also explored the role that depressive symptoms play in a model with alcohol use and alcohol-related negative consequences. The prevalence of college students who were diagnosed with depression in the last school year presents a great need to study its relationship with these constructs. Students with poor mental health or depression are also more likely to experience alcohol-related negative consequences (Weitzman, 2004), and there is a direct association between depressive symptoms and negative consequences, but not necessarily between alcohol use and depressive symptoms (Vickers, Patten, Bronars, Lane, Stevens, Croghan, Schroeder, & Clark, 2004). One study also suggests that protective behaviors partially mediate the relationship between depressive symptoms and negative consequences (Martens, Martin, Hatchett, Fowler, Fleming, Karakashian, & Cimini, 2008). Further, students with depressive symptoms who use protective behaviors drink less and experience fewer negative consequences, as compared to students without depressive symptoms who use protective behaviors (LaBrie, Kenney, Lac, Garcia, & Ferraiolo, 2009). Sleeping problems play an important role in the relationship between alcohol consumption and alcohol-related negative consequences. Poorer global sleep quality is associated with alcohol-related negative consequences after controlling for alcohol use. Further, among heavier drinkers, those with poorer sleep quality experienced greater levels of negative consequences than those who had better sleep quality (Kenney, LaBrie, Hummer, & Pham, 2012). The purpose of this study was to examine the relationships between alcohol use measured by estimated Blood Alcohol Content (eBAC), PBS, depression, and sleep problems, as they explain the variance of alcohol-related negative consequences using the spring 2009 national aggregate data set of the American College Health Association National College Health Assessment (ACHA-NCHA). This dataset was comprised of a random sample of undergraduate and graduate students from 117 U.S. colleges and universities (n=53,850). Reliability analyses, confirmatory factor analysis (CFA) and structural equation modeling (SEM) were used for model specification and evaluation. Model fit indices for the current study indicate that the model and the data in this study are a good fit, demonstrated by RMSEA= .044, 90% CI (.044, .044) and SRMR= .066. Findings suggest that an additive effect of eBAC, PBS, depression, sleep problems, and certain demographics explain 39% of the variance in alcohol-related negative consequences and greatly impact the amount of harm that college students may experience as a result of their alcohol use. Results from the current study may assist clinicians and health educators who want to improve the probability that they will be able help reduce negative consequences among college students when they drink alcohol. These staff may engage students in a conversation about risk reduction (e.g. one on one consults, campus-wide media campaign) and also provide support for conducting brief screenings about alcohol so that clinicians may be more effective in helping students to reduce alcohol-related negative consequences. The results from this study may also assist researchers in finding more relationships that account for some of the unexplained variance in this study. Interpreting these predictive relationships are important to the way that students are screened for alcohol problems on college campuses, as well as decisions that college students make about alcohol in the greater context of healthy lifestyle decisions. Future research could include repeating the analysis with each race/ethnicity separated out instead of as a dichotomous variable (white/non-white), conducting a similar analysis with each negative consequence instead of as a scale, developing a more complete sleep problems scale within the ACHA-NCHA with improved reliability, and a further investigation into the positive correlation between sleep problems and depression in order to explore other variables that mediate the relationship between depression and sleep problems among college students.

Chronic obstructive pulmonary disease (COPD) has a severe impact on quality of life (QOL). Using the Health Promotion Model as a guide, a cross-sectional, correlational design was used to describe relationships among individual characteristics and experiences (age, gender, race, severity of illness, resilience), behavior-specific cognitions and affect (self-efficacy, barriers, social support), behavioral outcomes (health promoting lifestyle), and QOL in this patient population. One hundred and twenty participants were recruited from three clinics at Virginia Commonwealth University Health System. In addition to a demographic survey, participants completed a 151-item questionnaire incorporating measures resilience, severity of illness, self-efficacy, and barriers to a health-promoting lifestyle, social support, lifestyle, and QOL. Spirometric evaluation of lung function and the 6-minute walking test were also completed. Structural equation modeling was used to determine the effect of nine independent variables on QOL.Participants were white (51.2%), female (63.6%), and approximately 60.5 years old. Severity of illness, characterized by symptoms and functional capacity, suggested they were not severely ill (mean = 3.18, S.D.= 2.69). They were somewhat resilient (mean = 136.01, S.D.= 23.01), had adequate social support (mean = 68.10, S.D.= 19.95), were uncertain about their competency (self-efficacy) to manage their health (mean = 24.91, S.D.= 4.92), sometimes experienced barriers (mean = 33.33, S.D.= 9.02), and sometimes included attributes of a healthy lifestyle in their lives (mean = 123.93, S.D.= 25.22). Their QOL was fair to poor (mean = 6.10, S.D.= 2.39).A series of analyses using structural equation modeling was conducted. The first model that was tested did not fit the data χ2(df = 13)= 67.989,p = 0.000, GFI = 0.895, CFI = 0.781, RMSEA = 0.189). Next, modification indices were use to reexamine for fit. Using the recommended modifications, a good fit model was obtained χ2(df = 9)=5.016, p = 0.833, GFI = 0.992, CFI = 1.0, RMSEA = 0.0); however, non-significant paths were present. An alternative model was tested and fit the data very well χ2(df=18)= 10.011, p = 0.932, GFI = 0.981, CFI = 1.0, RMSEA = 0.0). The independent variables explained about 45.1% of the variance in health-promoting lifestyle. All the variables explained 45.3% of variance in QOL. The most significant predictor of a healthy lifestyle was social support (0.383) and the most significant predictor of QOL was self-efficacy (0.364). The findings confirmed the utility of the HPM.

Event history analysis is the most prevalent modeling technique used to model event occurrence with longitudinal data (Cox & Oakes, 1984; Menard, 1991; Singer & Willett, 1993, 2003). An alternative is to model longitudinal data within the SEM framework, known as latent variable growth modeling (McArdle, 1988; Meredith & Tisak, 1990), which can provide a more robust framework. Whether or not a student remains in college presents an appropriate context within which to examine the modeling of event occurrence with longitudinal data. The purpose of the study was to compare event history and latent growth modeling as for predicting change in college student perseverance, with college student persistence literature serving as the framework. Students are defined as having persevered if they have earned hours and the end of the semester rather than if they are enrolled at the beginning of the semester, which is the traditional definition of persistence.The population for the study was the 2001 and 2002 cohorts of first-time, full-time freshmen at a large mid-Atlantic urban research university. Stopouts and transfer students were excluded. Data was analyzed for the first five semesters for each cohort. The results showed that parameter estimates were quite consistent across model type and time frame and were mostly consistent with previous research. No one method outperformed the others in terms of predicting correct classification. Using event history analysis with the structural equation modeling framework, however, appeared to be a very promising alternative to event history analysis with logistic regression since one can model error term and examine the differential effects of predictors at each time period. Finally, while latent growth modeling did not outperform the other methods in predictive classification, the study demonstrated it can be used for event occurrence analysis to test more complex theories.

The purpose of this study was to assess the impact of acquiescence on both positively and negatively worded questions, both when unidimensionality was assumed and when it was not. To accomplish this, undergraduate student responses to a previously validated survey of student engagement were used to compare several models of acquiescence, using a priori goodness-offit statistics as evidence for model fit, in order to develop a model that adequately accounted for acquiescence bias. Using a true experimental design, undergraduate students from a variety of classes at a large, urban university were randomly assigned to one of three versions of the same survey of student engagement (all positively worded items, all negatively worded items, an equal balance of both positively and negatively worded items). Structural equation modeling was used to analyze the results. Although the presence of acquiescence was confirmed for both positively and negatively worded items, it was not consistent by content scale or item polarization. This suggests that there may be an interaction between item polarization and content that may cause acquiescence to be present or absent. The scales that did not show acquiescence on the balanced survey portrayed a split factor loading based upon item polarization. Further, the splitting of factor loadings by item polarization was not due to acquiescence, suggesting that something other than acquiescence is causing the loadings to split. Further research is needed to develop models and/or methods to better assess and control for acquiescence. Although demographic groups were compared by gender and race/ethnicity to assess if different groups acquiesced differently, using multi-group confirmatory factor analysis, many of the models did not converge. The findings of this study were limited by the nature of the sample size. Additional research is needed to determine if acquiescence differs by group membership.

In spite of the wealth of research in IS security, there is very little understanding of what actually makes an IS security program successful within an organization. Success has been treated generally as a separate entity from IS security altogether; a great deal of research has been conducted on the "means to the end", while limited research has been focused on truly understanding what the end actually is. The problem compelling this research is that previous studies within the IS security domain do not adequately consider what factors contribute towards IS security success within the organizational context, and how the factors interact. This study built upon Shannon and Weaver (1949) and Mason (1978) to develop a model for predicting IS security success within an organization. A considerable body of information systems security literature was organized based on their findings. Core dimensions of information system security success were identified and operationalized within a model for predicting success with IS security initiatives. The model was empirically validated in a three-phase approach using survey methodology. First, the survey was tested for validity and reliability using an expert panel and pilot study. Next, the survey was administered to a sample, with the results analyzed using Confirmatory Factor Analysis and Structural Equation Modeling techniques. Initial analysis of the measurement model generated through Confirmatory Factor Analysis showed mixed fit. Factor loadings and average variance extracted calculations resulted in the selection of low performing items for removal; after revision, the revised measurement model showed improved fit for all measures. Structural Equation Modeling analysis was conducted on three structural models with varying levels of mediation. Based on the analysis of fit and comparison indices, the model depicting partial mediation was determined to be the best variation of the IS security success model. This study is the first known instance of an empirically tested IS security success model and should provide many avenues for future study, as well as providing practitioners a fundamental roadmap for success within their organizational IS security programs.

Structural equation models combine factor analysis models and multivariate regression models to estimate associations between observed variables and unobserved variables. The main achievement of this Capstone Project is the understanding of structural equation models and application of the models to real-world data. In this report, we reviewed structural equation models and several prerequisite topics. We performed a simulation study to compare maximum likelihood structural equation model estimation versus two-stage sequential estimation using multiple linear regression and maximum likelihood factor analysis. The simulation study confirmed that confidence intervals produced by structural equation models are valid and those obtained by two-stage sequential estimation are largely inaccurate. We applied structural equation models to an educational data comparing the efficacy of teaching conditions on learning scientific inquiry skills among 177 middle school students in Fitchburg, Massachusetts using a computer simulated science microworld. Application of structural equation models to the educational data showed that there were no significant differences in test score gains between three learning conditions, while controlling for latent factors measured by survey responses.

Dissertation submitted to the Department of Psychology, University of the Witwatersrand, in fulfilment of the requirements for the degree of Master of Arts. / The aim of the present study was to develop and evaluate a process model linking stressful environmental change, perceived job insecurity and stress reactions, A review of the literature revealed that stressful changes in the environment are linked to individual stress reactions, Stress theory has also recognised that appraisal of a stressful situation leads to stress reactions, Moreover, the literature has identified job Insecurity as a form of appraisal in that it is an internal event reflecting a transformation of beliefs about what 's happening in the organisation and environment. Job insecurity in turn has been shown to result in various stress reactions in individuals, Based on such research and theorising, a causal model was developed and tested using structural equation modeling techniques, It was assessed whether: stressful environmental change impacted upon stress reactions and job insecurity; job insecurity impacted upon stress reactions; and whether Job insecurity operated as a form of appraisal in mediating the relationship between stressful environmental change and stress reactions, The Independent variable, stressful environmental change, was specified as a common factor of the measured variables, political change, social change and organisational change, The proposed mediator variable, job insecurity, was specified as a common factor of the measured variables perceived threat to total job multiplied by powerlessness, and perceived threat to job features multiplied by powerlessness, The dependent variable, stress reactions, was specified as a common factor of the measured variables psychological distress, job dissatisfaction and reduced organisational commitment. The model was tested empirically using a combined sample of 267 subjects from three organisations, Results indicated that all relationships In the proposed model were confirmed, and that a reasonable fit was demonstrated between the empirical data and the theoretical model. Stressful environmental change was causally related to both stress reactions and Job insecurity, Job Insecurity was causally related to stress reactions, and in addition operated as a partial mediator between stressful environmental change and stress reacdons. Conceptual and methodological reasons for the findings are discussed, as well as some theoretical and practical implications, Limitations in the methodology are identified and future considerations of research are suggested, / AC2017