Dependencies in multilevel models at the second level have never been thoroughly examined. For certain designs first-level subjects are independent over time, but the second level subjects may exhibit nonzero covariances over time. Following a review of revelant literature the first study investigated which widely used computer programs adequately take into account these dependencies in their analysis. This was accomplished through a simulation study with SAS, and examples of analyses with Mplus and LISREL. The second study investigated the impact of two different missing data techniques for such designs in the case where data is missing at the first level with a simulation study in SAS. The first study simulated data produced in a multiyear study varying the numbers of subjects in the first and second levels, the number of data waves, the magnitude of effects at both the first and second level, and the magnitude of the second level covariance. Results showed that SAS and the MULTILEV component in LISREL analyze such data well while Mplus does not. The second study compared two missing data techniques in the presence of a second level dependency, multiple imputation (MI) and full information maximum likelihood (FIML). They were compared in a SAS simulation study in which the data was simulated with all the factors of the first study and the addition of missing data varied in amounts and patterns (missing completely at random or missing at random). Results showed that FIML is superior to MI because it produces lower bias and correctly estimates standard errors

One of the concerns in the field of statistics is the presence of missing data, which leads to bias in parameter estimation and inaccurate results. However, the multiple imputation procedure is a remedy for handling missing data. This study looked at the best multiple imputation methods used to handle mixed variable datasets with different sample sizes and variability along with different levels of missingness. The study employed the predictive mean matching, classification and regression trees, and the random forest imputation methods. For each dataset, the multiple regression parameter estimates for the complete datasets were compared to the multiple regression parameter estimates found with the imputed dataset. The results showed that the random forest imputation method was the best for mostly a sample of 150 and 500 irrespective of the variability. The classification and regression tree imputation methods worked best mostly on sample of 30 irrespective of the variability.

Thesis advisor: Lillie L. Albert / Increasing students' intentions to pursue mathematics-intensive careers is an urgent priority in the United States. To foster these intentions among marginalized student groups, such as immigrant students, and achieve equity in their career options, a critical question is whether we should allocate a greater proportion of school resources to mathematics over other subjects. The aims of this dissertation study were, first, to conceptually model and statistically evaluate how a school environment that prioritizes mathematics over other subjects might influence students' intentions to pursue mathematics over other academic subjects in the long term, and second, how this relationship is mediated by students’ mathematics pursuit attitudes, subjective norms, and perceived behavior (Ajzen, 1991), and moderated by their immigrant standing. The data for this study stemmed from the U.S. 2012 Programme For International Student Assessment Academic & Science (PISA) Student Questionnaire and School Questionnaire. A predictive mean matching technique was used to impute missing data that would resemble observed data. A 2-1-1 multilevel mediation Structural Equation Modeling (SEM) was implemented to accurately measure a school-level effect and student-level effect of the relationship of the examined constructs and to test the hypothesized model for the total sample. In order to compare immigrant student group and non-immigrant student group in the path model, multiple group path analysis was conducted. The results of the multilevel SEM model for the total sample presented that, at the school level (level 2), the school’s mathematics resources had no statistically significant direct and indirect effects on aggregated students’ intentions to pursue mathematics over other subjects. However, at the student level (level 1), students’ experiential and instrumental attitudes toward the pursuit of mathematics were positively related to students’ intentions to pursue mathematics over other subjects. The results of the multiple group path analysis comparing immigrant and non-immigrant student groups also found that the school’s mathematics resources had no statistically significant direct and indirect effects on students’ intentions to pursue mathematics over other subjects. However, a statistical difference in the overall path model of these two groups was found. The implications of this study for researchers, educators, and policymakers were discussed. / Thesis (PhD) — Boston College, 2021. / Submitted to: Boston College. Lynch School of Education. / Discipline: Teacher Education, Special Education, Curriculum and Instruction.

Les troubles musculo-squelettiques (TMS) sont la principale cause d’absentéisme au travail. Ils représentent un coût économique important, mais ils ont aussi de graves conséquences au niveau individuel – douleurs persistantes, limitations fonctionnelles ou perte d’emploi. De nombreuses expositions professionnelles ont déjà été mises en évidence pour expliquer la présence de ces symptômes, mais certaines relations restent encore à confirmer.L’objectif de ce travail de thèse était d’étudier l’association entre les facteurs professionnels, en particulier les facteurs physiques, et les TMS au niveau des coudes et des genoux. Les TMS au niveau des coudes ont été étudiés à travers la prévalence des symptômes aux coudes et des épicondylites, et l’incidence des épicondylites. Les douleurs aux genoux ont été étudiées à travers l’incidence des douleurs de courte ou de longue durée. Ce travail a nécessité l’utilisation de méthodologies spécifiques, en particulier les diagrammes causaux et les méthodes de gestion des données manquantes qui seront explicitées dans une partie théorique. La population dans ce travail de thèse était une cohorte de 3 710 sujets représentatifs des actifs des Pays de la Loire. Ils ont été recrutés entre 2002 et 2005 et ont été suivis entre 2007 et 2010. A chaque phase, un questionnaire sur les conditions de travail a été rempli par les participants et un examen clinique qui évaluait la présence de troubles musculo-squelettiques a été réalisé par des médecins du travail volontaires. En conclusion, les facteurs professionnels mis en évidence étaient globalement des facteurs de mouvements répétitifs impliquant les articulations étudiées (torsion des poignets et flexion/extension des coudes ou le fait de s’agenouiller). Le travail en force (manipulation de charges, efforts physiques importants) et les tâches répétitives ont également été mis en évidence comme prédicteurs de TMS ultérieurs / Musculoskeletal disorders (MSDs) are a leading cause of absenteeism from work. In addition to their major economic impact, musculoskeletal disorders have important consequences on individuals with the persistence of pain, disability and potential job loss. Many occupational exposures were highlighted to explain these symptoms, but certain relationships between occupational exposures and MSDs should be examined further.The aim of this thesis is to investigate the association between occupational exposures, especially biomechanical exposures, and elbow or knee MSDs. The elbow MSDs are studied with the prevalence of symptoms and lateral epicondylitis, and the incidence of lateral epicondylitis. The incidence of knee pain is examined according to its duration.The methodologies and theories appropriate to this type of work is briefly presented - a section focuses on the causal diagrams and methods of handling missing data.The population used is a cohort of 3710 subjects representative of the French workforce. Subjects were selected from workers undergoing a mandatory annual health examination between 2002 and 2005 and were followed between 2007 and 2010. For each phase of the study, a self-administered questionnaire on working conditions was completed and a clinical examination was performed in order to evaluate the presence of musculoskeletal disorders. In conclusion, repetitive movements involving the joints studied such as wrist-twisting, flexion/extension of the elbows or kneeling are associated with MSDs. The notion of force, such as load handling and physical exertion, and the notion of repetitive tasks are also identified as predictors of subsequent MSDs