The table lists each factor and interaction for generic error rates. Extended Data Table 5-1 lists the models and associated probabilities, used to compute the effects. P(incl) is the prior probability of the predictors; P(incl|d) is the posterior probability; BF_incl is Bayes factor. A BF greater than 1.0 favors the predictor, a BF less than 1.0 favors a null effect instead. The only predictors with favorable evidence, albeit of very modest entity, are block and probe dimension.