摘要：To examine the transfer of function phenomenon two 3-member equivalence classes were established using the typicalselection response of key pressing (A 1, 81, C1, & A2, 82, C2).One 3-member class was then made functionally equivalent bytraining a topographically distinct response (clapping) to A 1 andobtaining clapping at 81 and C1. Five experiments examined theeffects of adding a new function (stamping) to members of thisoriginal functional equivalence class. In Experiment 1 a newfunction was added to a three-member functional equivalenceclass by training it to a stimulus other than where the originalfunction trained. The results indicated that for adults and 12-yearoldchildren the new function transferred to all original classmembers. However, for two 6-yr-old children the new function didnot transfer to all class members. Experiments 2, 3, 4, and 5 wereconducted exclusively with children who were approximately 6years old and addressed issues arising from the findings ofExperiment 1. Specifically, these experiments investigated theeffects of the number of presentations of the stimuli in the transferof function testing stage, where the new function was trained inrelation to the original function, when the function was trained inrelation to the actual equivalence training, and how the functionswere trained. The general finding was that the new function rarelytransferred to all members of an established three-memberfunctional equivalence class for children who were approximately6 years old, unless the new function was trained to the stimuluswhere the original function was trained. These findings may haverelevance to the debate on both the constitution of classmembership and contextual control of equivalence responding.
