ANOVA and t-tests are used to obtain p-values, which are then used to calculate the chance of observing a significant mean difference. From a Bayesian standpoint, this is so-called frequentist method of determining a large difference in means is considered fundamentally incorrect, particularly in some cases it is evident that the null hypothesis is false. However, a Bayesian approach to inference for the age and race data appears to be doomed mainly because, although the data exist, we do not know of any previous distributions, have any idea how those prior distributions might look, or have any idea in what sense the age and race data is evidence of change.