|
| Relations Among Stimuli in Equivalence Classes |
| Tuesday, June 1, 2004 |
| 12:00 PM–1:20 PM |
| Fairfax A |
| Area: EAB; Domain: Applied Research |
| Chair: Lanny Fields (Queens College/SUNY) |
| Abstract: . |
| |
| The Nodal Structure of Equivalence Classes: Background |
| LANNY FIELDS (Queens College/CUNY), Mari Watanabe-Rose (Queens College/CUNY) |
| Abstract: The emergence of an equivalence class is defined by the mutual selection of stimuli in a set of N stimuli after the establishment of –1 relations by training. These test performances have been taken to mean that all of the stimuli in the class are functionally interchangeable and thus must be equally related to each other regardless of the number of nodes that separated the stimuli during training. A number of experiments have shown, however, that many test performances during delayed emergence, and during post class formation tests of various sorts are inverse functions of the number of nodes that separated the stimuli during training. These so called nodal distance effects are contrary the view that all of the stimuli in an equivalence class are equally related to each other. None of these studies, however, have attempted to use nodal distance as an independent variable that can precisely predict and control the bifurcation of a previously established equivalence class. Such a demonstration would strengthen the notion that the nodal structure of an equivalence class can be a strong determinant of the membership of stimuli in an equivalence class and would further the debate about the role played by nodal structure among the stimuli in equivalence classes. |
| |
| Effects of Nodal Structure on the Bifurcation of Equivalence Classes |
| MARI WATANABE-ROSE (Queens College/CUNY), Lanny Fields (Queens College/CUNY) |
| Abstract: Two 4-node 6-member equivalence classes emerged after training AB, BC, CD, DE, and EF. Then subjects were taught to make different responses to the C1 and C2 stimuli by discrimination training. A transfer test that involved the presentation of all 12 stimuli in the two classes showed immediate transfer of the C1 response two the other members of Class 1, and of the C2 response to the other members of Class 2. These data showed that the 4-node 6-member equivalence classes were also acting as 4-node 6-member functional classes. Thereafter, new responses were trained to the D1 and D2 stimuli by discrimination training. When the transfer test was repeated, the response trained to C stimulus in a class was evoked by the B and A stimuli from the same class, while the response trained to the D stimulus was evoked by the E and F stimuli from the same class. These data demonstrated that the 4-node 6-member equivalence/functional class was bifurcated into two 3-member functional classes in accordance to the nodal structure of the class. In addition, these data support the view that the relations among the stimuli in an equivalence class cannot be characterized in terms of interchangeability. Immediately after the C/D transfer test, subjects were re-exposed to the initial emergent relations test and responded in a manner that demonstrated the intactness of the original 4-node 6-member equivalence class. Thus, the bifurcation of the 6-member classes did not interfere with the integrity of the original 4-node 6-member classes. |
| |
| Stimulus Control During the Acquisition of Conditional Discriminations: A Kernel Analysis |
| MICHELLE C. GARRUTO (Queens College/CUNY), Lanny Fields (Queens College/CUNY), Mari Watanabe-Rose (Queens College/CUNY) |
| Abstract: When two-choice conditional discriminations are trained, there are 16 different forms of stimulus control that can govern behavior on a moment to moment basis. Each can be tracked by use of a kernel analysis of responding. When training was conducted to establish the five linked conditional discriminations AB, BC, CD, DE, and EF, responding was controlled by comparison position, comparison preferences, discriminative control of positional responding by the sample stimuli, inverse conditional relations between the sample and negative comparisons and eventually the conditional relation between the samples and positive comparisons. These data show that intermediate trial based percentage correct measured of stimulus control do not adequately characterize the processes involved in the formation of sets of linked conditional discriminations. Differ forms of preliminary training produced different patterns of stimulus control topographies that influenced performances during the acquisition of conditional discriminations. |
|
| |