Talk:Cognition August2018/@comment-35453425-20180528203135/@comment-35481611-20180530131702

Yeah we're definitely on the same page. I have found it tremendously difficult to evaluate any of the "intuitive statistician" papers because of their inattention to more rote associative principles. The authors of the parrot paper we read last week were at least conscious of them, even if the design still begot matching-law-like responding.

Why the Alonso-Diaz & Cantlon would be innovative were it to be applied to nonhumans is that responding is based on a dimension other than just the quantity estimation. Use of the total number of samples (cardinality, in this paper) as additional, statistically-meaningful information would I think be closer to that intuitive statistician bar.