UNREINFORCED CONDITIONAL SELECTION WITHIN 3-CHOICE CONDITIONAL DISCRIMINATIONS

Five normally developing 7- to 12-year-old children were taught two, 3 -choice conditional discriminations [A-C (i.e., A(1)-C-1, A(2)-C-2, an d A(3)-C-3) and B-C] and then given tests for equivalence (A-B and B-A ). This was repeated with two new conditional discriminations (D-F and E-F). The tests showed the development of six, 3-member equivalence c lasses. Subjects were then presented a new conditional discrimination with the ''D'' stimuli as samples and the ''C'' stimuli as comparisons without differential consequences. Three of the five subjects respond ed conditionally, that is, they selected a different comparison stimul us in the presence of each sample stimulus. Subsequent tests showed th at the six stimulus classes merged into three classes. The fourth subj ect responded inconsistently on the unreinforced D-C trials, but after additional conditional discrimination training and subsequent testing under a partial reinforcement baseline, she responded conditionally w ith respect to two of the three sample stimuli. The fifth subject sele cted the same ''C'' stimulus in the presence of all ''D'' stimuli. The unreinforced conditional selection shown here appears to be one examp le of generalized conditional responding. Other examples include learn ing set and exclusion.