A Bayesian Solution to Non-convergence of Crossed Random Effects Models

Crossed random effects models can simultaneously take into account both fixed and of the subjects stimuli when observations are nested within combinations stimuli. Unfortunately, maximum likelihood estimation (MLE) restricted (REML) often encounter convergence problems, which in turn lead researchers to fit simpler that yield invalid statistical inferences. On other hand, if structure is too simple, tests not valid; effect complex, inefficient. This study examines non-convergence issues inherent with MLE REML as well whether using Bayesian solve problems crossed models.