|
|
Type of Document Master's Thesis Author Rausch, Joseph R Author's Email Address jrausch@nd.edu URN etd-07092004-131544 Title DESIGNING LONGITUDINAL STUDIES OF NEGATIVE EXPONENTIAL GROWTH ACCORDING TO THE RELIABILITIES OF GROWTH PARAMETER ESTIMATORS Degree Master of Arts Department Psychology Advisory Committee
Advisor Name Title Scott E. Maxwell Committee Member Keywords
- nonlinear mixed models
- nonlinear multilevel models
- half-life
- parameterization
- nonlinear least squares
- nonlinear growth
- reliability
Date of Defense 2004-06-23 Availability restricted Abstract Researchers in the social and behavioral sciences frequently depend on linear models when designingand analyzing longitudinal studies. However, when attempting to precisely understand the underlying
growth process, nonlinear growth models are typically more useful. The present work focuses on a relatively simple nonlinear growth model, the negative exponential growth model. In particular, a new parameterization of this growth model is proposed to facilitate meaningful interpretation of its parameters. Further, one of the primary goals of the present work is to develop a method that allows researchers to design longitudinal studies according to the population reliabilities of the nonlinear least square estimators of the negative exponential growth parameters. The method that is recommended makes use of Monte Carlo simulation to approximate the population reliabilities in a
relatively accurate manner. An illustration of this method is presented and discussed to promote
the design of efficient longitudinal studies of nonlinear growth in the behavioral sciences.
Files
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access RauschJR072004.pdf 1.00 Mb 00:04:38 00:02:23 00:02:05 00:01:02 00:00:05 indicates that a file or directory is accessible from the campus network only.