Multilevel models are one of the most frequently used methods for analyzing multilevel data. These types of data occur when observations (Level 1) are clustered within higher-level collectives (Level 2), for example, students nested in schools or employees nested in work teams. Unfortunately, multilevel data often contain missing data, for example, when participants omit certain items in a questionnaire or they drop out before the end of a study. If treated improperly, missing data can distort parameter estimates and compromise statistical (...) decision making. For this reason, it is often recommended to use principled methods for dealing with missing data such as multiple imputation (MI) or maximum likelihood estimation (ML). In the present dissertation, I consider the treatment of missing data with an emphasis on multilevel MI. In multilevel research, it is important that the imputation model takes the multilevel structure into account. However, many open questions remain about how this can be achieved in practice. In multiple studies, I considered different implementations of multilevel MI and examined how missing data can be treated in multilevel models with random intercepts, random slopes, interaction effects, continuous and categorical data, and missing data at Level 2. In addition, I considered the analysis of multiply imputed data sets and examined different procedures for pooling the results of multiparameter tests (e.g., model comparisons), including applications in the analysis of variance (ANOVA) and multilevel models with tests about fixed effects and variance components. Finally, I present the R package mitml, which provides a set of practical tools for conducting multilevel MI. This includes tools for the specification of the imputation model, convergence diagnostics, analyzing multiply imputed data sets, and various pooling methods along with a tutorial article that illustrates its features and provides a nontechnical introduction to multilevel MI. (shrink)

In the past 2 decades latent variable modeling has become a standard tool in the social sciences. In the same time period, traditional linear structural equation models have been extended to include non-linear interaction and quadratic effects (e.g., Klein and Moosbrugger, 2000 ), and multilevel modeling (Rabe-Hesketh et al., 2004 ). We present a general non-linear multilevel structural equation mixture model (GNM-SEMM) that combines recent semiparametric non-linear structural equation models (Kelava and Nagengast, 2012 ; Kelava et al., 2014 ) with (...) multilevel structural equation mixture models (Muthén and Asparouhov, 2009 ) for clustered and non-normally distributed data. The proposed approach allows for semiparametric relationships at the within and at the between levels. We present examples from the educational science to illustrate different submodels from the general framework. (shrink)

Conventionally, multilevel analysis of covariance (ML-ANCOVA) has been the recommended approach for analyzing treatment effects in quasi-experimental multilevel designs with treatment application at the cluster-level. In this paper, we introduce the generalized ML-ANCOVA with linear effect functions that identifies average and conditional treatment effects in the presence of treatment-covariate interactions. We show how the generalized ML-ANCOVA model can be estimated with multigroup multilevel structural equation models that offer considerable advantages compared to traditional ML-ANCOVA. The proposed model takes into account measurement (...) error in the covariates, sampling error in contextual covariates, treatment-covariate interactions, and stochastic predictors. We illustrate the implementation of ML-ANCOVA with an example from educational effectiveness research where we estimate average and conditional effects of early transition to secondary schooling on reading comprehension. (shrink)

Technical best management practices are the dominant approach promoted to mitigate agriculture’s significant contributions to environmental degradation. Yet very few social science studies have examined how farmers actually use these practices. This study focuses on the outcomes of farmers’ technical best management practice adoption related to synthetic nitrogen fertilizer management in the context of Midwestern corn agriculture in the United States. Moving beyond predicting the adoption of nitrogen best management practices, I use structural equation modeling and data from a sample (...) of over 2500 farmers to analyze how the number of growing season applications a farmer uses influences the rate at which synthetic nitrogen is applied at the field-level. I find that each additional application of N during the growing season is associated with an average increase of 2.4 kg/ha in farmers’ average N application rate. This result counters expectation for the outcome of this practice and may suggest that structural pressures are leading farmers to use additional growing season applications to ensure sufficiently high N rates, rather than allowing them to reduce rates. I conclude by discussing the implication of this study for future research and policy. (shrink)

This dissertation is a defense of a particular theory of the metaphysics of time which I call "accretivism", but which is popularly known in a form usually called the "Growing Block Theory". The goal of a metaphysics of time is to incorporate the various aspects of our temporal experience into a single, comprehensive whole. To this end I delineate five aspects of our ordinary experience of time: 1) The Tensed Aspect, in virtue of which objects are presented to us as (...) past, present, or future; 2) The Transitory Aspect, in virtue of which time passes or "flows"; 3) The Durational Aspect, in virtue of which entities have a certain temporal extent; 4) The Structural Aspect, in virtue of which entities are given as being in temporal relations to one another, and 5) The Differential-Repetitive Aspect, in virtue of which things are different from one time to another, and yet there is a certain recurrence of aspects of our experience form one time to another. I contrast the accretivist picture of time, according to which that which is past and that which is present both have ontological status, but nothing which is future has ontological status, and in which temporal passage consists in the coming-into-being of new entities at the temporal edge of reality marked by the present, with the two dominant theories of time in the contemporary literature: 1) presentism, according to which only that which is present has ontological status, and 2) four-dimensionalism, according to which time is to be understood on analogy with spatial dimensions. Accretivism, I argue, is superior to the other two views in virtue of the fact that it gives full status to both the Structural Aspect of Time, for which the presentist has difficulty accounting, and the Transitory Aspect of Time, for which the four-dimensionalist has difficulty accounting. I then defend the accretivist picture against a variety of objections that might be raised to it. (shrink)