There has been a wealth of research conducted on the high school dropouts spanning several decades. It is estimated that compared with those who complete high school, the average high school dropout costs the economy approximately $250,000 more over his or her lifetime in terms of lower tax contributions, higher reliance on Medicaid and Medicare, higher rates of criminal activity, and higher reliance on welfare (Levin & Belfield, 2007). The nation suffers not only because of the loss in revenue but also as a result of the education level of the population. Individuals who choose to drop out of high school are less likely to be in the labor force than adults who earned a high school credential, and they fare worse in many aspects of life. In many studies on high school dropouts, an important challenge is how to determine an appropriate structural form for a statistical model to be used in making inferences and predictions. Many useful statistical modeling for survival analysis have been developed to study the competing risks frame of probability of dropping out and the probability of graduating; however, few methods exist for establishing the actual competing risks structural form of a model when the data contains two educational milestones – drop out and graduation. In this dissertation, we first utilized the data collected from the National Education Longitudinal Study (NELS: 88/2000) and proposed a discrete time competing risks hazard model and the corresponding model selection process to study the contributions of student’s academic ability, family background, school characteristics and vocational education to the probabilities of students graduating from or dropping out of high school. This model finds a way to overcome the shortcomings of the traditional models existing in the previous research. Within educational research, missing data is very common occurrence and can easily complicate the model selection problem. Handling missing data inappropriately can lead to bias and inaccurate inferences. This dissertation applies four missing data techniques to the key attributes including listwise deletion, dummy variable adjustment, mean imputation, and multiple imputation. Recommendations were offered for future endeavors and research in finding solutions to handle missing data in educational research. Finally, we outline the implementation of the proposed methodology. This research has the potential for both theoretical merit and implications for affecting educational policy. My dissertation adds to the limited body of literature of quantitative studies of the high school dropouts. A discrete time competing risks hazard model for predicting the probability of dropping out could become part of a powerful tool to identify students at risk of dropping out.

Assessing risk had always been the key focus when it comes to recidivism. Using risk assessment instruments, it is possible to predict the outcome of recidivism dichotomously. These measures, however, can only predict between 70-80 percent of validity, and they specify only levels of risk (low-medium-high), but not time. Therefore, the aim of this study is to define time of recidivism to supplement risk assessment with a possible new actuarial approach and fill out gaps in the existing literature. To do so a systematic literature review was conducted with a controlled search on exact time points. All the fourteen studies resulted in the final model were: published in the past 20 years, had some connection to time and were quantitative. The results highlighted the importance of the first year, especially the first half of the year as the most critical period regarding recidivism. Three different time periods were identified: (1) the end of the first month; (2) between the second and the third; (3) at the end of the 6th. Consequently, it might be beneficial to target these critical periods with more intense supervision/intervention in order to decrease the likelihood of recidivism.

In longitudinal cohort studies, potential risk factors are measured at baseline, subjects are followed over time, and disease endpoints are ascertained via extensive surveillance. Individual follow-up time is from baseline to the event, if one is observed during the study period. Follow-up time is censored for subjects who are not observed to have the event during the study period, at the end of the study period for subjects who remain event-free, but during the study period for subjects who leave the study early by choice or by mortality, or whose last evaluation was before the end of the study. Survival analytic techniques are unique in that the unit of analysis is not the individual but the person-time contributed by the individual. Surveillance in longitudinal studies is generally quite rigorous. Subjects are examined in waves and their event status is ascertained. Surveillance continues between waves, and events come to the attention of the investigator. If there is a long time between waves, analyses can be conducted on all available data, with non-events censored early at the last examination and events followed beyond the general examination to the incident event. Motivated by analyses using the Framingham Heart Study (FHS) with cardiovascular endpoints, we consider four censoring methods for non-events and evaluate their impact on estimates of incidence, and on tests of association between risk factors and incidence. We further investigate the impact of early censoring of non-events (as compared to events) under various scenarios with respect to incidence estimation, robustness, and power using a simulation study of Weibull survival models over a range of sample sizes and distribution parameters. Our FHS and simulation investigations show early censoring of non-events causes over estimation of incidence, particularly when the baseline incidence is low. Early censoring of non-events did not affect the robustness of the Wald test [Ho: Hazard Ratio (HR) =1]. However, in both the FHS and over the range of simulation scenarios, under early censoring of non-events, estimates of HR were closer to the null (1.0), and the power to detect associations with risk factors was markedly reduced.

Widespread critical shortages of high-quality teachers in the United States (Sutcher, Darling-Hammond, Carver-Thomas, 2016) has prompted considerable research on staffing trends within the teaching profession. Research suggests both an increase in the demand for teachers and a "chronic and relatively high annual turnover compared with many other occupations" (Ingersoll & Smith, 2003, p. 31). Recent studies have highlighted the negative effects that high teacher turnover has on financial costs, school climate, and student performance. Since attrition rates appear to be higher for beginning teachers (Ingersoll & Smith, 2003; Ingersoll, 2012), it is important to understand why beginning teacher attrition occurs and what factors influence beginning teachers to stay in the profession, move to another school, or return to the profession. While several studies suggest multiple factors influence teacher attrition, having a better understanding of how these factors correlate with each other and how the impact of these factors changes over time will provide additional information into how time influences teacher attrition. Exploring where teaching go after they leave teaching and why some teachers decide to return to the profession will provide additional insight into the complex nature of teacher attrition patterns in the United States. The purpose of this study was to examine attrition patterns among K-12 teachers who began teaching in a public school in the United States during the 2007-2008 academic year and factors that influenced teachers decisions to move from their initial school to another school, discontinue teaching, or return to the position of a K-12 teacher. This study used data collected as part of the Beginning Teacher Longitudinal Study (BTLS) and explores the effect that various predictor variables have on the probability that BTLS teachers will either leave teaching or move to another school. Using a structural equation modeling (SEM) approach to discrete-time survival analysis made it possible to simultaneously model systems of equations that included both latent and observed variables, allow for the effect of mediators, and analyze how the effect of each predictor variable changed over time. Results suggest the higher the teachers' base salary during their first three years of teaching, the less likely they were to leave the profession during their second through fourth years of teaching. Teachers who supplement their base salaries by working extra jobs are more likely to leave the profession after their fourth year of teaching. Teachers who participated in an induction program during their first year of teaching were less likely to leave the profession in Wave 2 of the study and teachers who had taken more courses on teaching methods and strategies before they started teaching were less likely to leave teaching in all waves of the study than teachers who had taken fewer courses on teaching methods or strategies. Teachers who reported higher levels of positive school climate during their first year of teaching were less likely to leave the profession in Wave 2 and 4. Teachers who indicated higher levels of satisfaction with being a teacher in their school were less likely to move schools than teachers with lower levels of satisfaction and teachers who taught in schools with higher percentages of students who were approved for free or reduced prices lunches were more likely to move schools than teachers with lower percentages of students who were approved for free or reduced price lunches. However, due to convergence issues, these results should be interpreted with caution. Weighted item response descriptive analyses suggest teachers' most important reason for moving schools was to work in a school more convenient to their home. Teachers who leave teaching are more likely to enter professions or occupations in education-related fields than occupations outside the field of education. Results also suggest teachers who leave the profession of teaching are more likely to be working in a job, either full-time or part-time, than not working in job. Finally, the majority of teachers who return to the profession of teaching do so because they missed being a K-12 teacher or they want to make a difference in the lives of others. This study contributes to the existing literature on teacher attrition by testing whether multiple relationships exist between various predictor variables and beginning teacher attrition and examines how the influence of each of these predictor variables changes over time. The study also investigates topics that have been relatively unexplored in the literature, including where teachers go when they leave the profession and factors that influence teachers' decisions to return to the profession. The results of this study may benefit researchers, teachers, educators, administrators, and policy makers interested in and/or studying teacher attrition in the United States.

Background: Knowledge of the incidence of tuberculosis (TB) and hospitalization postpartum could reduce maternal morbidity and mortality. TB infections are prevalent in pregnant women living with Human immunodeficiency virus (HIV) compared to women not living with HIV in South Africa. Adherence to Antiretroviral Therapy (ART) is poor among pregnant and postpartum women living with HIV (WLHIV), thus making WLHIV at a higher risk of hospitalization postpartum, due to the increased risk of Cesarean delivery (CD) and obstetric conditions as a result of HIV. The prevalence of TB among pregnant and postpartum women is poorly defined including in high prevalence TB and HIV locations, indicating limited evidence. The aim is to explore the incidence of TB and hospitalization within four years postpartum among WLHIV, including associated risk factors. Methodology: The study population is from phase 2 of the Maternal and Child HealthAntiretroviral Therapy (MCH-ART) study. It is a single-arm observational cohort study of 628 WLHIV who attended antenatal care (ANC). Enrolment into phase 1 began in March 2013, the final deliveries from phase 2 were in December 2014, and the final follow-up visits were completed in 2016. MCH-ART is an ongoing study with global approval examining strategies for providing HIV care and treatment to HIV-infected women who initiate ART during pregnancy and their HIV-exposed infants. This study took place at the Midwife-Obstetric Unit (MOU) at Gugulethu Community Health Centre, Western Cape South Africa. It consists of three connected study designs and three phases through the antenatal and postnatal periods. Phase 1 is a cross-sectional study, phase 2 is a cohort study and phase 3 is a randomized trial. Kaplan-Meier survival analysis was used to assess the incidence of TB and hospitalization over time among ix WLHIV up to four years postpartum and Cox regression was used to measure the effect of risk factors on the incidence of TB and hospitalization. Results: Thirty-five (35) WLHIV developed TB postpartum at a total person-time of 2365.1 woman-years. The incidence rate (IR) of developing TB among WLHIV postpartum was 1.48 (95% CI=1.03-2.06) cases per 100 woman-years from 2013 to 2018. Twenty-three (23) WLHIV was hospitalized postpartum and a total person-time of 552.8 woman-years was spent. The IR of hospitalization among WLHIV postpartum was 4.16 (95% CI=2.64-6.24) cases per 100 womanyears from 2013 to 2018. The IR of TB and hospitalization among WLHIV postpartum is statistically significant. Adjusting, for other risk factors, the history of diabetes at ANC, the history of TB at ANC and CD4 count (200 - <500) cells/mm3 at ANC also significantly increases the incidence of TB postpartum, whereas, obstetric reasons is associated with the hospitalization of WLHIV.

Indiana University-Purdue University Indianapolis (IUPUI) / Survival analysis often encounters the situations of correlated multiple events including the same type of event observed from siblings or multiple events experienced by the same individual. In this dissertation, we focus on the joint modeling of bivariate time to event data with the estimation of the association parameters and also in the situation of a semi-competing risk. This dissertation contains three related topics on bivariate time to event mod els. The first topic is on estimating the cross ratio which is an association parameter between bivariate survival functions. One advantage of using cross-ratio as a depen dence measure is that it has an attractive hazard ratio interpretation by comparing two groups of interest. We compare the parametric, a two-stage semiparametric and a nonparametric approaches in simulation studies to evaluate the estimation perfor mance among the three estimation approaches. The second part is on semiparametric models of univariate time to event with a semi-competing risk. The third part is on semiparametric models of bivariate time to event with semi-competing risks. A frailty-based model framework was used to accommodate potential correlations among the multiple event times. We propose two estimation approaches. The first approach is a two stage semiparametric method where cumulative baseline hazards were estimated by nonparametric methods first and used in the likelihood function. The second approach is a penalized partial likelihood approach. Simulation studies were conducted to compare the estimation accuracy between the proposed approaches. Data from an elderly cohort were used to examine factors associated with times to multiple diseases and considering death as a semi-competing risk.

Indiana University-Purdue University Indianapolis (IUPUI) / In studies with competing risks outcomes, misidentifying the event-type responsible for the observed failure is, by definition, an act of misclassification. Several authors have established that such misclassification can bias competing risks statistical analyses, and have proposed statistical remedies to aid correct modeling. Generally, these rely on adjusting the estimation process using information about outcome misclassification, but invariably assume that outcome misclassification is non-differential among study subjects regardless of their individual characteristics. In addition, current methods tend to adjust for the misclassification within a semi-parametric framework of modeling competing risks data. Building on the existing literature, in this dissertation, we explore the parametric modeling of competing risks data in the presence of outcome misclassification, be it differential or non-differential. Specifically, we develop parametric pseudo-likelihood-based approaches for modeling cause-specific hazards while adjusting for misclassification information that is obtained either through data internal or external to the current study (respectively, internal or external-validation sampling). Data from either type of validation sampling are used to model predictive values or misclassification probabilities, which, in turn, are used to adjust the cause-specific hazard models. We show that the resulting pseudo-likelihood estimates are consistent and asymptotically normal, and verify these theoretical properties using simulation studies. Lastly, we illustrate the proposed methods using data from a study involving people living with HIV/AIDS (PLWH)in the East-African consortium of the International Epidemiologic Databases for the Evaluation of HIV/AIDS (IeDEA EA). In this example, death is frequently misclassified as disengagement from care as many deaths go unreported to health facilities caring for these patients. In this application, we model the cause-specific hazards of death and disengagement from care among PLWH after they initiate anti-retroviral treatment, while adjusting for death misclassification. / 2021-03-10

Indiana University-Purdue University Indianapolis (IUPUI) / The crossing of Kaplan-Meier survival curves presents a challenge when conducting survival analysis studies, making it unclear whether any of the study groups involved present any significant difference in survival. An approach involving the determination of maximum vertical distance between the curves is considered here as a method to assess whether a survival advantage exists between different groups of patients. The method is illustrated on a dataset containing survival times of patients treated with two cancer treatment regimes, one involving treatment by chemotherapy alone, and the other by treatment with both chemotherapy and radiotherapy.

No description available.

In this research, the fire effects on structural and compositional change, and advance regeneration of oak forests in the Ozarks of Missouri were investigated by combining the statistic methods of MANONA, survival analysis, CART analysis, and logistic analysis. Results indicated that fire treatments significantly reduced the midsotry and understory basal area and stem density. However, fire effects on overstory tree survival differentiated among size classes. A new morphological variable, ratio of the total height to the square of basal diameter, was found to be statistically significantly related to the tree mortality rate for most of the species. The developed logistic regression models for selected species using the morphological variable well simulated the impact of initial stem size of advance regeneration on mortality for most of the species. The resultant logistic regression models could be a potential tool to compare and quantify species response to fires on a comparable basis.