8 1st graders, 13 retardates, and 4 monkeys were given 2-choice discrimination problems composed of 3 objects: P, always rewarded; N, always nonrewarded; and A, ambiguously rewarded depending upon whether paired with P or N on any given trial. A prompting procedure controlled instrumental responses to the rewarded object (P on PA trials and A on NA trials) and presumably equated the development of independent approach tendencies to P and A. 3 levels of prompting (0, 8, or 16 prompted trials) preceded conventional test trials. Results, similar for all Ss, suggested that Ss learned about both objects on each trial and that response tendencies were developed to all 3 objects independently. (PsycINFO Database Record (c) 2006 APA, all rights reserved)