Variables affecting the formation of stimulus-classes: Studios on the effects of training arrangement

Abstract

The present essay describes studies analysing effects of training designs with a single node and with multiple nodes, on equivalence class formation and transfer of discriminative functions. This analysis suggests that training design and number of nodes do not have an absolute effect, but interact in a complex way with other procedural parameters. Some studies show that designs with a single node are more effective than designs with multiple nodes. However, modifications in the procedure of conditional discrimination training and in the topography of selection responses may tender designs with multiple nodes as effective as designs with a single node. lt is suggested that these variations are related to stimulus control topographies in baseline conditional discriminations induced by different aspects of the training procedures.