|
Activity Number:
|
405
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
WNAR
|
| Abstract - #301508 |
|
Title:
|
Capturing Group Membership via Growth Mixture Models: A Simulation Study
|
|
Author(s):
|
Kevin Delucchi*+ and Alan Bostrom
|
|
Companies:
|
University of California, San Francisco and University of California, San Francisco
|
|
Address:
|
401 Parnassus Ave., San Francisco, CA, 94143,
|
|
Keywords:
|
latent variable models ; longitudinal ; growth mixture ; simulation
|
|
Abstract:
|
Results of a series of simulations examining the ability of latent growth mixture models to capture population group membership, intercept and slope parameters is presented. The model was a common design; longitudinal with four equally-spaced assessments and two latent groups. 56 conditions were simulated from a 7 x 2 x 2 x 2 factorial design with 1000 samples per condition for total Ns of 300 and 900 with sample size imbalance ranging from small to large. Percent correctly classified ranged from 58% to 88%. Estimates of population slope and intercept ranged from acceptable to poor depending on conditions simulated. Additional simulations to be reported indicate large sample sizes are needed for this type of modeling. The addition of an informative covariate improved the estimates.
|
