The importance of isomorphism for conclusions about homology: A Bayesian multilevel structural equation modeling approach with ordinal indicators

Supplementary files.

Guenole, N. (in review). The importance of isomorphism for conclusions about homology. A bayesian multilevel structural equation modeling approach with ordinal indicators.

Description:  Van de Schoot et al. (2015) observed that variance parameters estimated with Bayesian methods can be subject to spikes (i.e., extreme estimates) especially for variance terms, which inflate parameter estimates. To permit readers to evaluate whether this occurred in the current Monte Carlo study this file contains trace plots for the within and between latent variance and latent residual variance parameters for a sample run from each of the 384 cells in the design. Each file is numbered according to the design cell represents and contains a panel plot of four figures a) within latent variance, b) within latent residual variance, c) between latent variance, and d) between latent residual variance.

References

Van de Schoot, R., Broere, J.J., Perryck, K., Zondervan-Zwijnenburg, M., & Van Loey, N. (2015). Analyzing Small Data Sets using Bayesian Estimation: The case of posttraumatic stress symptoms following mechanical ventilation in burn survivors. European Journal of Psychotraumatology, 6: 25216 - http://dx.doi.org/10.3402/ejpt.v6.25216