The Impact of Inaccurate “Informative” Priors for Growth Parameters in Bayesian Growth Mixture Modeling
Author:
- 3 April 2014
- journal article
- research article
- Published by Taylor & Francis in Structural Equation Modeling: A Multidisciplinary Journal
- Vol. 21 (2) , 239-252
Abstract
Within Bayesian estimation, prior distributions are placed on model parameters and these distributions can take on many different levels of informativeness. Although much of the research conducted within this estimation framework uses what are called diffuse (or noninformative) priors, there are certain models and modeling circumstances where it is more optimal to use what are referred to as informative priors. This study focuses on the latter situation and examines the effects of inaccurate informative priors on the growth parameters within the context of growth mixture modeling. Overall, results indicated that growth mixture modeling is relatively robust to the use of inaccurate mean hyperparameters for the growth parameters, as long as the variance hyperparameters are somewhat large.Keywords
This publication has 14 references indexed in Scilit:
- Evaluation of Structural Equation Mixture Models: Parameter Estimates and Correct Class AssignmentStructural Equation Modeling: A Multidisciplinary Journal, 2010
- Nonlinear Growth Models in Mplusand SASStructural Equation Modeling: A Multidisciplinary Journal, 2009
- Bayesian analysis of longitudinal data using growth curve modelsInternational Journal of Behavioral Development, 2007
- How vague is vague? A simulation study of the impact of the use of vague prior distributions in MCMC using WinBUGSStatistics in Medicine, 2005
- Probabilistic Sensitivity Analysis of Complex Models: A Bayesian ApproachJournal of the Royal Statistical Society Series B: Statistical Methodology, 2004
- Methodological Advances in the Analysis of Individual Growth With Relevance to Education PolicyPeabody Journal of Education, 2002
- 10.1007/978-1-4899-4485-6_1Crossref Listing of Deleted Dois, 2000
- Eliciting expert beliefs in substantial practical applications [Read before The Royal Statistical Society at ameeting on 'Elicitation' on Wednesday, april 16th, 1997, the President, Professor A. F. M. Smithin the Chair]Journal of the Royal Statistical Society: Series D (The Statistician), 1998
- The Selection of Prior Distributions by Formal RulesJournal of the American Statistical Association, 1996
- Explaining the Gibbs SamplerThe American Statistician, 1992