Effects of response bias on learning and memory tasks in squirrel monkeys

Abstract

Six squirrel monkeys were tested on short-term memory tasks assessing ability to suppress perseverative responses that had been previously reinforced. Each trial was divided into three parts: Initial Preference Assay (IPA), Bias-Conditioning (BC), and Reversal Conditioning (RC), and alternated between two conditions: experimental and control. Strength of response bias (based on choices of response during IPA) exceeded chance levels for each monkey. Eighty-four percent of responses to BC of the experimental trials were made to the response loci chosen in IPA even though that response was not rewarded. Monkeys made 38% correct responses during RC but shifted from making most errors during control trials in the beginning of the experiment to making most errors during experimental trials by the end of the experiment. Monkeys had developed a strategy of persevering from IPA to BC and then shifting to the other, not previously chosen window on RC, which led to correct responses in the experimental trials. (Abstract shortened with permission of author.)