Fig. 6From: Using an onset-anchored Bayesian hierarchical model to improve predictions for amyotrophic lateral sclerosis disease progressionThe effect of changing the number of months of data used for prediction in both the linear and onset-anchored models. This effect is measured by the MSE (and associated 90% credible interval) resulting from a single replicate of the cross-validation analysis for both models. As the length of data-capture increases the benefits from including an anchor decreasesBack to article page