How experimental evidence shifts the probability assigned to each competing explanation in a contrast set.