Fig. 2: Bayesian analysis of cognitive tests in the ADNI dataset.
From: Identifying healthy individuals with Alzheimer’s disease neuroimaging phenotypes in the UK Biobank
These distributions represent the Bayesian posterior estimates of the mean of cognitive tests in Alzheimer’s disease (AD) score positive and negative groups, with the Region Of Practical Equivalence (ROPE) as a shaded column. As expected in this well-characterised dataset, for all measures there was very strong evidence of a difference between groups classified as positive or negative by AD score derived from structural neuroimaging, indicated by mean AD score in the AD score positive group and the 95% credible intervals (indicated by the thin horizontal bars) falling outside the ROPE. The mean of the AD score negative group is represented by the dotted vertical line with the ROPE denoted by the shaded area on each side. The 75% credible interval is denoted by the thick bars and the 95% credible interval by the thin bars.
