1,721,248 research outputs found
Marginal longitudinal nonparametric regression: locality and efficiency of spline and kernel methods
No full textWe consider nonparametric regression in a longitudinal marginal model of generalized estimating equation (GEE) type with a time-varying covariate in the situation where the number of observations per subject is finite and the number of subjects is large. In such models, the basic shape of the regression function is affected only by the covariate values and not otherwise by the ordering of the observations. Two methods of estimating the nonparametric function can be considered: kernel methods and spline methods. Recently, surprising evidence has emerged suggesting that for kernel methods previously proposed in the literature, it is generally asymptotically preferable to ignore the correlation structure in our marginal model and instead assume that the data are independent, that is, working independence in the GEE jargon. As seen through equivalent kernel results, in univariate independent data problems splines and kernels have similar behavior; smoothing splines are equivalent to kernel regression with a specific higher-order kernel, and hence smoothing splines are local. This equivalence suggests that in our marginal model, working independence might be preferable for spline methods. Our results suggest the opposite; via theoretical and numerical calculations, we provide evidence suggesting that for our marginal model, marginal smoothing and penalized regression splines are not local in their behavior. In contrast to the kernel results, our evidence suggests that when using spline methods, it is worthwhile to account for the correlation structure. Our results also suggest that spline methods appear to be more efficient than the previously proposed kernel methods for our marginal model
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Inference in frailty measurement error models.
No full textWe propose a new class of models, frailty measurement error models (FMEMs), for clustered survival data when covariates are measured with error. We explore FMEMs from three directions: bias analysis, structural modeling and functional modeling. We study the asymptotic bias when measurement error is ignored and when the underlying distribution of the unobserved error-prone covariates is misspecified. We found that ignoring measurement error in covariates will underestimate estimates of regression coefficients and overestimate variance components. As the censoring proportion increases, the attenuation of estimation of regression coefficients becomes more severe. However, it is not necessarily true for the variance component estimation. Structural modeling and functional modeling is developed to make statistical inference in FMEMs. Under structural modeling, we assume a distribution for the unobserved error-prone covariates and calculate nonparametric maximum likelihood estimates (NPMLEs) using an EM algorithm. Under functional modeling, we make no distributional assumption on the unobserved error-prone covariates and use the SIMEX method for parameter estimation. The NPMLEs and SIMEX estimates are compared in terms of efficiency and robustness. NPMLE gives more efficient estimators when correctly specifying the distribution of unobserved covariates X, and could yield biased estimates when such distribution is misspecified. The SIMEX approach is model robust with respect to the misspecification of the distribution of X. However, it yields less efficient estimates than NPMLE and is computationally intensive. We use the SIMEX approach to test the variance components in FMEMs and extended the results to the discrete frailty measurement error models. All the proposed methods are applied to the west Kenya parasitemia data and their performance is evaluated through simulations.PhDMathematicsPure SciencesStatisticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/132186/2/9959807.pd
Latent variable models for longitudinal data with multiple outcomes, informative dropouts and missing covariates.
No full textIn many studies the outcome of main interest cannot be measured by a single response. There is a great deal of literature dealing with such data for cross-sectional studies. However, this problem has not been well studied for longitudinal data. In this dissertation we propose latent variable models to handle this type of multivariate longitudinal data. At the first stage of the model we assume the observed outcomes measure the latent variable with error. The latent variable is then assumed associated with covariates through a linear mixed model. We extend this model to the situation where the probability of dropout is latent variable dependent, and hence non-ignorable. We first show how one can find maximum likelihood estimates when the covariates are completely observed. We then relax this assumption by allowing covariates to be missing due to unit dropout, which is often the case when there are time-varying covariates. Finally, we look at the missing covariate issue in more detail for the single outcome case. We carry out a bias analysis, comparing our proposed method with naive methods for handling the missing covariates. The Gibbs sampler for this model is developed to obtain Bayesian inference. Data from a national panel study on changes in methadone treatment practices are used throughout to illustrate the methodology.PhDBiological SciencesBiostatisticsPure SciencesStatisticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/132674/2/9977251.pd
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Bayesian inference in generalized additive mixed models.
No full textWe propose Bayesian generalized additive mixed models for correlated data, which arise frequently in studies involving clustered, hierarchical and spatial designs. The models allow for additive functional dependence of a continuous or discrete outcome variable on covariates by using nonparametric regression and account for correlation between observations using random effects. Partially improper integrated Wiener priors are used for the nonparametric functions and the resulting estimators are cubic smoothing splines. When the distribution of the random effects is normal, a Gibbs sampling algorithm is provided for the estimation of all model parameters and inference for fixed effects, random effects, and nonparametric functions. Systematic inference can be made within a modified generalized linear mixed model framework. We also propose a generalized additive mixed model which relaxes the normality assumption for the distribution of the random effects. A Dirichlet process prior distribution is assumed for the random effects. Computation is carried out using Gibbs sampling. Systematic inference on model parameters can be made within a modified generalized linear mixed model framework without parametric distributional assumption on random effects. We illustrate the proposed approaches by analyzing two real-world data sets and evaluate their performance through simulations.PhDBiological SciencesBiostatisticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/124122/2/3121988.pd
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Transition measurement error models for longitudinal data.
No full textWe propose a new class of models, transition measurement error models, to model longitudinal data when covariates are measured with error. We study the asymptotic bias in the estimation when the measurement error is ignored. Such bias can be large. We propose three approaches for inference in transition measurement error models to account for the measurement error. These approaches include the structural modeling approach, the simulation-extrapolation (SIMEX) approach, and the semi-parametric estimation approach. When a structural model is correctly specified for the unobserved true covariate, the maximum likelihood estimates for the regression parameters in the transition models are derived via the Expectation-Maximization algorithm. When no knowledge is available about the distribution of the unobserved true covariate, both the SIMEX approach and the semiparametric estimation approach are proposed: the SIMEX approach is simulation-based and the semiparametric estimation approach is likelihood-based. We show that the estimators using the SIMEX approach are consistent and asymptotically normal with a known correct extrapolation function. To implement the semiparametric estimation approach, we construct the conditional score equations to give consistent estimators for the parameters in both linear transition model and logistic transition model. These estimators are also shown to be asymptotically normal. Specifically, we are able to obtain the most efficient estimator for linear transition models in the presence of validation data. All three approaches are applied to a longitudinal social support study for elderly women with heart disease.PhDBiological SciencesBiostatisticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/123255/2/3068941.pd
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
