Parsimonious Mixed Models
arXiv:1506.04967
Abstract
The analysis of experimental data with mixed-effects models requires decisions about the specification of the appropriate random-effects structure. Recently, Barr, Levy, Scheepers, and Tily, 2013 recommended fitting `maximal' models with all possible random effect components included. Estimation of maximal models, however, may not converge. We show that failure to converge typically is not due to a suboptimal estimation algorithm, but is a consequence of attempting to fit a model that is too complex to be properly supported by the data, irrespective of whether estimation is based on maximum likelihood or on Bayesian hierarchical modeling with uninformative or weakly informative priors. Importantly, even under convergence, overparameterization may lead to uninterpretable models. We provide diagnostic tools for detecting overparameterization and guiding model simplification.
ArXiv preprint. 21 pages, 6 figures
Cited by in corpus (11)
- Balancing Type I Error and Power in Linear Mixed Models
- Statistical methods for linguistic research: Foundational Ideas - Part II
- Bayesian linear mixed models using Stan: A tutorial for psychologists, linguists, and cognitive scientists
- Bayesian Item Response Modeling in R with brms and Stan
- Generalised additive mixed models for dynamic analysis in linguistics: a practical introduction
- Statistical methods for linguistic research: Foundational Ideas - Part I
- Mixed Effects Models are Sometimes Terrible
- Score-based Tests for Explaining Upper-Level Heterogeneity in Linear Mixed Models
- Choosing the correlation structure of mixed effect models for experiments with stimuli
- Selection of inverse gamma and half-t priors for hierarchical models: sensitivity and recommendations
- Applied Regression Analysis of Correlations for Correlated Data