Matching Items (3)
Filtering by
- All Subjects: longitudinal data
- Creators: Grimm, Kevin
Description
This paper investigates a relatively new analysis method for longitudinal data in the framework of functional data analysis. This approach treats longitudinal data as so-called sparse functional data. The first section of the paper introduces functional data and the general ideas of functional data analysis. The second section discusses the analysis of longitudinal data in the context of functional data analysis, while considering the unique characteristics of longitudinal data such, in particular sparseness and missing data. The third section introduces functional mixed-effects models that can handle these unique characteristics of sparseness and missingness. The next section discusses a preliminary simulation study conducted to examine the performance of a functional mixed-effects model under various conditions. An extended simulation study was carried out to evaluate the estimation accuracy of a functional mixed-effects model. Specifically, the accuracy of the estimated trajectories was examined under various conditions including different types of missing data and varying levels of sparseness.
ContributorsWard, Kimberly l (Author) / Suk, Hye Won (Thesis advisor) / Aiken, Leona (Committee member) / Grimm, Kevin (Committee member) / Arizona State University (Publisher)
Created2016
Description
The first half-year of infancy represents a salient time in which emotion expression assumes a more psychological character as opposed to a predominantly physiological one. Although previous research has demonstrated the relations between early parenting and later emotional competencies, there has been less of a focus on differentiating positive and negative emotion expression across the early infancy period. Thus, the current study investigates the growth of positive and negative emotion expression across early infancy in a low-income, Mexican-American sample, and examines the development of emotion expression as a function of early maternal emotion socialization and prenatal stress. Participants included 322 mothers and their infants. Data were collected in participants' homes prenatally and when the infants were 12-, 18-, and 24-weeks old. Mothers were asked to interact with their infants in a semi-structured teaching task, and video-taped interactions of mother and infant behaviors were then coded. Data for mothers was collected at the prenatal and 12-week visits and data for infants was collected at the 12-, 18-, and 24-week visits. Prenatal stress was measured via two questionnaires (Daily Hassles Questionnaire and Perceived Stress Scale). Maternal socialization at 12 weeks was represented as a composite of four observational codes from the Coding Interactive Behavior coding system. Infant emotion expression was also globally rated across the 5-minute teaching task. Findings suggest that the normative development of emotion expression across early infancy is complex. Positive emotion expression may increase across the early infancy period whereas negative emotion expression decreases. Further, at 12 weeks, greater maternal emotion socialization relates to more infant positivity and less negativity, in line with current conceptualization of parenting. However, across time, greater early socialization predicted decreased positivity and was unrelated to negative emotion expression. Findings also suggest that prenatal stress does not relate to socialization efforts or to infant emotion expression. A better understanding of the nuanced development of positive and negative emotion development as a function of early parenting may have implications for early intervention and prevention in this high-risk population.
ContributorsRoss, Emily (Author) / Crnic, Keith (Thesis advisor) / Grimm, Kevin (Committee member) / Eisenberg, Nancy (Committee member) / Arizona State University (Publisher)
Created2015
Description
Longitudinal recursive partitioning (LRP) is a tree-based method for longitudinal data. It takes a sample of individuals that were each measured repeatedly across time, and it splits them based on a set of covariates such that individuals with similar trajectories become grouped together into nodes. LRP does this by fitting a mixed-effects model to each node every time that it becomes partitioned and extracting the deviance, which is the measure of node purity. LRP is implemented using the classification and regression tree algorithm, which suffers from a variable selection bias and does not guarantee reaching a global optimum. Additionally, fitting mixed-effects models to each potential split only to extract the deviance and discard the rest of the information is a computationally intensive procedure. Therefore, in this dissertation, I address the high computational demand, variable selection bias, and local optimum solution. I propose three approximation methods that reduce the computational demand of LRP, and at the same time, allow for a straightforward extension to recursive partitioning algorithms that do not have a variable selection bias and can reach the global optimum solution. In the three proposed approximations, a mixed-effects model is fit to the full data, and the growth curve coefficients for each individual are extracted. Then, (1) a principal component analysis is fit to the set of coefficients and the principal component score is extracted for each individual, (2) a one-factor model is fit to the coefficients and the factor score is extracted, or (3) the coefficients are summed. The three methods result in each individual having a single score that represents the growth curve trajectory. Therefore, now that the outcome is a single score for each individual, any tree-based method may be used for partitioning the data and group the individuals together. Once the individuals are assigned to their final nodes, a mixed-effects model is fit to each terminal node with the individuals belonging to it.
I conduct a simulation study, where I show that the approximation methods achieve the goals proposed while maintaining a similar level of out-of-sample prediction accuracy as LRP. I then illustrate and compare the methods using an applied data.
I conduct a simulation study, where I show that the approximation methods achieve the goals proposed while maintaining a similar level of out-of-sample prediction accuracy as LRP. I then illustrate and compare the methods using an applied data.
ContributorsStegmann, Gabriela (Author) / Grimm, Kevin (Thesis advisor) / Edwards, Michael (Committee member) / MacKinnon, David (Committee member) / McNeish, Daniel (Committee member) / Arizona State University (Publisher)
Created2019