Dissertation Information for Yeow Meng ThumNAME:
DEGREE:
DISCIPLINE:
SCHOOL: ADVISORS: COMMITTEE MEMBERS: MPACT Status: Incomplete - Not_Inspected Title: Analysis of individual variation: A multivariate hierarchical linear model for behavioral data Abstract: The theoretical role of individual differences in psychology has been a divisive issue among psychologists for a long time. It began as a fundamental philosophical disagreement but it quickly spawned two competing research strategies with respect to the role of individual difference variables in explaining behavior. The result is that neither approach, by itself, is able to give as complete an account of psychological functioning as researchers in each discipline have hoped. This research proposes a resolution to the conceptual impasse which involves a population concept of behavioral processes. To complement this new conceptual paradigm, a two-stage multivariate hierarchical linear model (MHLM) is developed. Although this model maintains a clear analytic separation between the behavioral fluctuations of an individual and the variation of individual performance in the population, it provides a description of the observable features of individual functioning within an integrated individual difference framework. Special attention is given to three common characteristics of behavioral data: the use of multiple outcomes, some outcomes are unobserved by design or by neglect, and the number of subjects may be small. We consider maximum likelihood estimation for the normal-normal model as well as for the normal-t model. The estimation procedure is based on a reliable variable-metric algorithm which supports linear equality constraints on point estimators. Estimation of the normal-t model involves numerical integration by a one-dimensional Gauss-Laguerre quadrature. Three computational examples illustrative of the procedure are given in addition to two applications which exemplify the two-stage approach in psychology--the estimation of generalized learning curves and the analysis of test-validation data. |
MPACT Scores for Yeow Meng ThumA = 0 Advisors and Advisees Graph |