Bayesian model criticism: prior sensitivity of the posterior predictive checks method

Use of noninformative priors with the Posterior Predictive Checks (PPC) method requires more attention. Previous research of the PPC has treated noninformative priors as always noninformative in relation to the likelihood, regardless of model-data fit. However, as model-data fit deteriorates, and the steepness of the likelihood's curvature diminishes, the prior can become more informative than initially intended. The objective of this dissertation was to investigate whether specification of the prior distribution has an effect on the conclusions drawn from the PPC method. Findings indicated that the choice of discrepancy measure is an important factor in the overall success of the method, and that different discrepancy measures are affected more than others by prior specification