Talk:Cognition August2018/@comment-35453425-20180530154508/@comment-35453425-20180530171021

I cannot figure out how to comment on your comments, so I will add here. I am not sayin they are NOT using probabilistic estimates. I actually think they might be, but there are alternate possibilities. Will's question to me is a good one. What we know is that the greater the ratio of small to large set, the slower the responding and the more errors that get made. We do not know if more bailing out would occur on those trials. HOWEVER, a paper we did some years ago, which has largely been ignored in the literature, shows that, across trials, chimps do pay attention to array sizes as they occur in choice tests, and then use that information in the face of ambiguity. Here is the link to that one: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2674346/

I would be keen to hear what they rest of you think about this paper of ours, and what it might add to this debate. In a nutshell, we saw that after making repeated judgments between two visible sets of foods, and then being asked to choose between a visible set and an unknown set, chimps used the approximate average reward from the first 15 trials (where both sets were seen) to decide whether to take the visible or take the unknown. Our key conclusion relevant to this debate was that "We believe that the chimpanzees attended to the outcomes of each trial and used trial-by-trial feedback of the quantity obtained to modify some representation of their overall reward rate for that day."