One single primary study is only a little piece of a bigger puzzle. Meta-analysis is the statistical combination of results from primary studies that address a similar question. The most general case is the random-effects model, in where it is assumed that for each study the vector of outcomes T_i~N(θ_i,Σ_i ) and that the vector of true-effects for each study is θ_i~N(θ,Ψ). Since each θ_i is a nuisance parameter, inferences are based on the marginal model T_i~N(θ,Σ_i+Ψ). The main goal of a meta-analysis is to obtain estimates of θ, the sampling error of this estimate and Ψ. Standard meta-analysis techniques assume that Σ_i is known and fixed, allowing the explicit modeling of its elements and the use of Generalized Least Squares as the method of estimation. Furthermore, one can construct the variance-covariance matrix of standard errors and build confidence intervals or ellipses for the vector of pooled estimates. In practice, each Σ_i is estimated from the data using a matrix function that depends on the unknown vector θ_i. Some alternative methods have been proposed in where explicit modeling of the elements of Σ_i is not needed. However, estimation of between-studies variability Ψ depends on the within-study variance Σ_i, as well as other factors, thus not modeling explicitly the elements of Σ_i and departure of a hierarchical structure has implications on the estimation of Ψ. In this dissertation, I develop an alternative model for random-effects meta-analysis based on the theory of hierarchical models. Motivated, primarily, by Hoaglin's article "We know less than we should about methods of meta-analysis", I take into consideration that each Σ_i is unknown and estimated by using a matrix function of the corresponding unknown vector θ_i. I propose an estimation method based on the Minimum Covariance Estimator and derive formulas for the expected marginal variance for two effect sizes, namely, Pearson's moment correlation and standardized mean difference. I show through simulation studies that the proposed model and estimation method give accurate results for both univariate and bivariate meta-analyses of these effect-sizes, and compare this new approach to the standard meta-analysis method.

The Borg Ratings of Perceived Exertion (RPE) Scale has proven to be a highly popular instrument in measuring the subjective responses of individuals to a given work or exercise task. Historically, the instrument was designed to correlate highly with the heart rates in young-to-middle-aged men performing various tasks. The body of literature, however, has revealed inconsistencies in the extent of just how strong the relationship is between ratings of perceived exertion and various physiological criterion variables, most notably, heart rate. In addition, most studies have invoked the question of whether the criterion-related validity coefficients derived from the relationship between ratings of perceived exertion and a specified physiological criterion variable are just as valid as those for which the Borg RPE Scale was originally performed. A meta-analysis, therefore, was undertaken to determine the magnitude of the relationship between ratings of perceived exertion scores and each of three commonly used physiological measures or criterion variables: heart rate, blood lactate, and oxygen uptake. Results show that by using Tests of Homogeneity for each physiological criterion variable, the observed sample size-weighted validity coefficients are heterogeneous. The median of the mean sample size-weighted validity coefficients is .574 for heart rate, .561 for blood lactate, and .480 for oxygen uptake. Each study in the meta-analysis was grouped by the study characteristics of subject gender, fitness level, RPE Scale, exercise type, exercise protocol, and study quality. For heart rate, the highest validity coefficients are those in which the subjects are highly fit, the exercise type is fairly unusual, such as swimming, and the subjects are required to maximally exert themselves. For blood lactate, the highest validity coefficients are for females, healthy-inactive subjects, the 15-point RPE Scale, treadmill use, and swimming. For oxygen uptake, the highest validity coefficients between ratings of perceived exertion and oxygen uptake are for swimming. In a meta-analysis of study effects, when the validity coefficients are analyzed by study, the resultant mean validity coefficients are only somewhat higher (ratings of perceived exertion and heart rate, .657; ratings of perceived exertion and blood lactate, .642; ratings of perceived exertion and oxygen uptake, .609) than those obtained using sample size-weighted validity coefficients. Finally, corrections for bias generally resulted in increased validity coefficients and decreased variances.

The continued study of asymptomatic Ebolavirus infection is necessary to develop a more complete understanding of Ebola transmission dynamics. This paper conducts a meta-analysis of eight studies that measure seroprevalence (the number of subjects that test positive for anti-Ebolavirus antibodies in their blood) in subjects with household exposure or known case-contact with Ebola, but that have shown no symptoms. In our two random effects Bayesian hierarchical models, we find estimated seroprevalences of 8.76% and 9.72%, significantly higher than the 3.3% found by a previous meta-analysis of these eight studies. We also produce a variation of this meta-analysis where we exclude two of the eight studies. In this model, we find an estimated seroprevalence of 4.4%, much lower than our first two Bayesian hierarchical models. We believe a random effects model more accurately reflects the heterogeneity between studies and thus asymptomatic Ebola is more seroprevalent than previously believed among subjects with household exposure or known case-contact. However, a strong conclusion cannot be reached on the seriousness of asymptomatic Ebola without an international testing standard and more data collection using this adopted standard.

This meta-analysis focuses on the willingness to work aspect of the RAW model of employability of Hogan et al. (2009), in relationship to career success. Willingness to work (W) can be defined as favorably disposed to work hard and take initiative at one’s job. The variables I used to structure the W are proactive personality, conscientiousness, work ethic, job involvement, adaptability, and ambition. I used the Hunter and Schmidt method to analyze the data applying a random effects model. All calculations were conducted in Excel. The overall sample consisted of 100 effect sizes (r) derived from 41 studies. The total sample size was 45,652. The individuals in these samples were from a wide range of backgrounds that included diverse samples of age, culture, and occupations. The results indicated a small to medium effect size for all variables. This outcome supports my hypotheses, concluding that willingness to work correlates with both objective and subjective career success. Because of small sample sizes (i.e., relatively few studies with usable moderator data), a moderator analysis was not conducted. Once sufficient studies have been published in this domain, future researchers could look into the possibility of moderators.

This meta-analysis investigated the difference in perceptions of health-related quality of life (HRQOL) among long-term early-stage breast cancer survivors (BCS). The comparison was between African American and European American women. Initial pilot searches suggested that enough studies existed for a meaningful meta-analysis of a BCS population at least 5 years post diagnosis. Only studies using the outcome measure HRQOL were included in the study; this yielded an initial sample of 212 study reports, with 56 reports entering the coding phase of the process. African American women were grossly underrepresented in this set of studies in comparison to the overall breast cancer population. Separate analyses of Medical Outcomes Study 36- Item Short- Form Health Survey, Quality of Life-Cancer Survivor and Quality of Life Index - Cancer Version III instruments were executed. However, no stringent comparison across instruments of the difference between the HRQOL of African American and European American women was possible. When African American women were included in the populations, researchers often did not report their data separately but rather included their data in an overall population and thus differences were masked. The data that were available, including qualitative studies for African American women, suggested that there was a lower perception of the quality of survival in some areas for African American women. These differences suggest the need for greater attention to the physical components of African American BCS. The results point to a need to improve African American participant recruitment in research and to use online databases as a results repository to improve data availability for analysis.

Com o desenvolvimento do sequenciamento em larga escala, novas tecnologias surgiram para auxiliar o estudo de sequências de ácidos nucleicos (DNA e cDNA); como consequência, o desenvolvimento de novas ferramentas para analisar o grande volume de dados gerados fez-se necessário. Entre essas novas tecnologias, uma, em particular, chamada Imunoprecipitação de Cromatina seguida de sequenciamento de DNA em larga escala ou CHIP-Seq, tem recebido muita atenção nos últimos anos. Esta tecnologia tornou-se um método usado amplamente para mapear sítios de ligação de proteínas de interesse no genoma. A análise de dados resultantes de experimentos de ChIP-Seq é desaadora porque o mapeamento das sequências no genoma apresenta diferentes formas de viés. Os métodos existentes usados para encontrar picos em dados de ChIP-Seq apresentam limitações relacionadas ao número de amostras de controle e tratamento usadas, e em relação à forma como essas amostras são combinadas. Nessa tese, mostramos que métodos baseados em testes estatísticos de hipótese tendem a encontrar um número muito maior de picos à medida que aumentamos o tamanho da amostra, o que os torna pouco conáveis para análise de um grande volume de dados. O presente estudo descreve um método estatístico Bayesiano, que utiliza meta-análise para encontrar sítios de ligação de proteínas de interesse no genoma resultante de experimentos de ChIPSeq. Esse métodos foi chamado Meta-Analysis Bayesian Approach ou MABayApp. Nós mostramos que o nosso método é robusto e pode ser utilizado com diferentes números de amostras de controle e tratamentos, assim como quando comparando amostras provenientes de diferentes tratamentos. / With the development of high-throughput sequencing, new technologies emerged for the study of nucleic acid sequences (DNA and cDNA) and as a consequence, the necessity for tools to analyse a great volume of data was made necessary. Among these new technologies, one in special Chromatin Immunoprecipitation followed by massive parallel DNA Sequencing, or ChIP-Seq, has been evidenced during the last years. This technology has become a widely used method to map locations of binding sites for a given protein in the genome. The analysis of data resulting from ChIP-Seq experiments is challenging since it can have dierent sources of bias during the sequencing and mapping of reads to the genome. Current methods used to nd peaks in this ChIP-Seq have limitations regarding the number of treatment and control samples used and on how these samples should be used together. In this thesis we show that since most of these methods are based on traditional statistical hypothesis tests, by increasing the sample size the number of peaks considered signicant changes considerably. This study describes a Bayesian statistical method using meta-analysis to discover binding sites of a protein of interest based on peaks of reads found in ChIP-Seq data. We call it Meta- Analysis Bayesian Approach or MABayApp. We show that our method is robust and can be used for dierent number of control and treatment samples, as well as when comparing samples under dierent treatments.

Cognitive ability is one of the most frequently investigated individual differences in management and psychology. Countless studies have demonstrated that tests measuring cognitive ability or intelligence predict a number of important real-world outcomes such as academic performance, vocational training performance, and job performance. Although the relationship between intelligence and real-world performance is well established, there is a lack of consensus among scholars with regard to how intelligence should be conceptualized and measured. Of the more traditional theories of intelligence, two perspectives are particularly dominant: the Cattell-Horn model of fluid and crystallized intelligence and the theory of General Cognitive Ability (GCA or g). Fluid ability (Gf) represents novel or abstract problem solving capability and is believed to have a physiological basis. In contrast, crystallized ability (Gc) is associated with learned or acculturated knowledge. Drawing on recent research in neuroscience, as well as research on past performance, the nature of work, and expert performance, I argue that compared to measures of fluid ability, crystallized ability measures should more strongly predict real-world criteria in the classroom as well as the workplace. This idea was meta-analytically examined using a large, diverse set of over 400 primary studies spanning the past 100 years. With regard to academic performance, measures of fluid ability were found to positively predict learning (as measured by grades). However, as hypothesized, crystallized ability measures were found to be superior predictors of academic performance compared to their fluid ability counterparts. This finding was true for both high school and college students. Likewise, similar patterns of results were observed with regard to both training performance and job performance. Again, crystallized ability measures were found to be better predictors of performance than fluid measures. This finding was consistent at the overall level of analysis as well as for medium complexity jobs. These findings have important implications for both intelligence theory and selection practice. Contemporary intelligence theory has placed great emphasis on the role of fluid ability, and some researchers have argued that Gf and g are essentially the same construct. However, the results of this study, which are based on criterion-related validities rather than factor-analytic evidence, demonstrate that Gc measures are superior predictors in comparison to Gf measures. This is contrary to what one would expect if Gf and g were indeed the same construct. Rather, the findings of this study are more consistent with General Cognitive Ability theory, which predicts that Gc indicators will be the best predictors of future learning and performance. Given that Gc measures demonstrate higher criterion-related validities than Gf measures, Gc measures are likely to be preferred for selection purposes. Further, Gf scores are known to decline with age while Gc scores remain relatively stable over the lifespan. Thus, when used for selection purposes, Gf tests may underpredict the performance of older workers. In contrast, research has shown that Gc measures are predictively unbiased. Additional implications for theory and practice are discussed, along with study limitations and opportunities for future research.

Osteoarthritis is a common degenerative disorder that affects joints. Despite recent therapeutic advances, osteoarthritis continues to be a challenging health problem, and elderly population is particularly at risk. Pain is the most unbearable symptom experienced by osteoarthritic patients. Currently, several pharmacological medications are available to manage osteoarthritic pain. Opioids, potent analgesics, have shown extraordinary ability to reduce intense pain in many osteoarthritic clinical trials. Although many clinical trials have investigated the efficacy and safety of opioids in osteoarthritic patients, there is an increased need for a study to integrate the reported outcomes and utilize them to achieve a better understanding of efficacy and safety profiles of opioids. Therefore, in our present study, efficacy, safety, and tolerability profiles of opioid compounds used to manage osteoarthritic pain were assessed and compared using a model-based meta-analysis (MBMA). To achieve our goal, a comprehensive database consisting of pain relief compounds with information on summary-level of efficacy over time, adverse events and dropout rates was compiled from multiple sources. MBMA was conducted using nonlinear mixed-effects modeling approach. The results showed that the selected models successfully captured the observed data, and primary efficacy endpoint estimations indicated that the ED50 of oxycodone, oxymorphone, and tramadol were 47, 84, and 247 mg per day, respectively. Efficacy-time course analysis showed that opioids had rapid time to efficacy onset, suggesting potential powerful pain relief effects. Also, it was found that gastrointestinal adverse events were the most opioid-associated and dose-dependent adverse effects. In addition, the analysis revealed that opioids are well-tolerable at low to moderate doses. The results presented here provided clinically meaningful insights into the efficacy and safety of oxycodone, oxymorphone, and tramadol. In addition, the presented framework analysis has a clinical impact on drug development where it can help in optimizing the dose of opioids to manage osteoarthritic pain, making precise key decisions for positioning of new drugs, and designing more efficient clinical trials.

The vast and complex wealth of information available to researchers often leads to a systematic review, which involves a detailed and comprehensive plan and search strategy with the goal of identifying, appraising, and synthesizing all relevant studies on a particular topic. A meta–analysis, conducted ideally as part of a comprehensive systematic review, statistically synthesizes evidence from multiple independent studies to produce one overall conclusion. The increasingly widespread use of meta–analysis has led to growing interest in meta–analytic methods for rare events and sparse data. Conventional approaches tend to perform very poorly in such settings. Recent work in this area has provided options for sparse data, but these are still often hampered when heterogeneity across the available studies differs based on treatment group. Heterogeneity arises when participants in a study are more correlated than participants across studies, often stemming from differences in the administration of the treatment, study design, or measurement of the outcome. We propose several new exact methods that accommodate this common contingency, providing more reliable statistical tests when such patterns on heterogeneity are observed. First, we develop a permutation–based approach that can also be used as a basis for computing exact confidence intervals when estimating the effect size. Second, we extend the permutation–based approach to the network meta–analysis setting. Third, we develop a new exact confidence distribution approach for effect size estimation. We show these new methods perform markedly better than traditional methods when events are rare, and heterogeneity is present.

The psychological impacts of infertility have been well documented in the literature, providing evidence to support that at least some women who confront infertility are at risk for heightened distress and depressive symptoms. In response to this accumulated evidence, it has been argued that psychoeducational interventions may provide an important component to the treatment of infertility. While several theoretical models postulate the effects of stress on infertility treatment outcomes, results of these investigations have led to conflicting conclusions. However, a synthesis of the accumulated data examining the effects of stress on ART treatment outcomes was nonexistent until the conduct of this study. Therefore, the purpose of this study was to investigate the impact of stress on ART treatment outcomes and to determine whether psychoeducational interventions mitigate the impact of stress experienced by women during an ART treatment program. Two hypotheses were tested: 1. Increased levels of stress will be associated with a lower likelihood of Assisted Reproductive Technology (ART) treatment success, and 2. Psychoeducational interventions will mitigate the effects of stress experienced during Assisted Reproductive Technology (ART) treatment. A meta-analysis analyzing the results for each hypothesis was tested through a hierarchical linear regression model. A total of 13 studies, representing 43 effect sizes, were included in the analysis investigating the relationship between stress and ART treatment outcomes. Results of the HLM regression model suggest that stress has a small negative association with ART treatment outcomes (d=0.2012, p< .05). The analysis investigating the relationship between psychoeducational interventions and stress included a total of 4 studies, representing 12 effect sizes. Empirical evidence gathered through this analysis revealed that the effect of psychoeducational interventions on the stress experienced by women participating in an ART treatment program were not statistically significant (d=0.3071, p>.05). However, because this analysis was based on such a small sample of studies, generalizations regarding the efficacy of psychoeducational interventions cannot be made. Therefore, research aimed at investigating the impacts of a variety of programs should continue in an effort to provide more conclusive information.