Mayday for Maybery: A reply to an invalid critique of the mixed model of linear-syllogistic reasoning



Murray Maybery's critique of the mixed model of linear-syllogistic reasoning as applied to both determinate and indeterminate items is invalid. First, the model correctly predicts that heterogeneous indeterminate linear syllogisms should be harder, not easier, than determinate ones. Second, contrary to Maybery's claim, my coding of the mismatch parameter was correct. Third, the operations I suggest for distinguishing indeterminate from determinate linear syllogisms do indeed apply to all items I (or Herbert Clark) modelled, although there do exist other problems that might have been modelled for which the operations would be insufficient to make the distinction. These points notwithstanding, the mixed model has in common with other cognitive models that it is an approximation to what subjects do, not an exact rendition. And as I have shown earlier, there are individual differences in subject strategies that are worth taking into account.