|Stimulus Equivalence: From Conceptual Issues to Translational Research|
|Monday, May 28, 2018|
|11:00 AM–12:50 PM |
|Marriott Marquis, San Diego Ballroom C|
|Area: EAB/EDC; Domain: Translational|
|Chair: Erik Arntzen (Oslo and Akershus University College)|
|Discussant: Paula Ribeiro Braga Kenyon (Trumpet Behavioral Health)|
The symposium will focus on different dimensions of behavior analysis within research on stimulus equivalence. Hence, conceptual, experimental, and translation research will be exemplified. The symposium will include papers from four different laboratories. The first paper by Dr. Vaidya will discuss the Sidman’s (2000) theory of stimulus equivalence and that all the features of equivalence class formation result from the merger of individual contingencies due common elements. This address will critically evaluate the DOE literature with human subjects to determine if a detailed examination of the empirical evidence will support or refute an interpretation cast in terms of Sidman’s theory of equivalence relations. The second paper by Drs. Steingrimsdottir and Arntzen study the role of including reflexivity trials in the test for emergent relations in two experiments. Experiment 1 employed a linear series training structure while Experiment 2 employed a many-to-one training structure. The results so far show a difference in equivalence class formation with respect to inclusion of reflexivity trials depending on the training structure. The third paper by Dr. Albright et al. study relatedness of stimuli in equivalence class members in college students. The main findings showed that relatedness was influenced by the combination of nodal number and relational type. Furthermore, the experiment showed a strong post class formation graded nodal distance effect. The fourth paper by Drs. Fields and Reeve is focusing on translation research. The paper will present protocols for the establishment of stimulus classes will be considered as vehicles to enhance the generalization to “novel exemplars” after the training of a target response. This approach is an example of translational research.
|Keyword(s): basic research, stimulus equivalence, translational research|
Can Sidman's Theory Provide an Alternative Interpretation of the Differential Outcomes Effect?
|MANISH VAIDYA (University of North Texas)|
The Differential Outcomes Effect (DOE) is usually attributed to the development of stimulus-specific outcome expectancies that supplement the control exerted by antecedent stimuli to facilitate learning or terminal performances. Sidman's (2000) theory of stimulus equivalence proposes that all elements of a contingency of reinforcement become part of an equivalence relation and that the traditional relations of symmetry, transitivity, and equivalence result from the merger of individual contingencies due to common elements. The theory also requires that common elements that conflict with the development of reinforced conditional relations selectively drop out of the equivalence relation and implicitly predicts that contingencies that avoid class merger would produce faster acquisition relative to contingencies which do not avoid the class merger. These predictions are routinely borne out in the DOE literature. This address will critically evaluate the DOE literature with human subjects to determine if a detailed examination of the empirical evidence will support or refute an interpretation cast in terms of Sidman's theory of equivalence relations.
|The Inclusion of Reflexivity Test Trials in Stimulus Equivalence Research|
|HANNA STEINUNN STEINGRIMSDOTTIR (Oslo and Akershus University College), Erik Arntzen (Oslo and Akershus University College)|
|Abstract: The three defining properties of the equivalence relation are reflexivity, symmetry and transitivity. However, the reflexivity test trials are often omitted in stimulus equivalence research. The purpose of the current experiments was to inspect the effect of including reflexivity trials during equivalence testing through different manipulations of the training and test trials. In Experiment 1, we used a linear-series training structure for training 12 conditional discriminations and testing for three 5-member equivalence classes (see upper part of Table 1). The main findings showed that if classes were not established, the participants did not respond in accordance with reflexivity. In Experiment 2, we used a many-to-one training structure. The preliminary results showed that the participants were most likely to respond in accordance with reflexivity when equivalence classes were formed (see lower part of Table 1). The results will be discussed with suggestions to further studies.|
|Nodality and the Graded Strengths of Transitive and Equivalence Relations|
|Leif Albright (Caldwell University), Lanny Fields (Queens College, City University of New York), KENNETH REEVE (Caldwell University), Sharon A. Reeve (Caldwell University), April N. Kisamore (Caldwell University)|
|Abstract: Prior research has found that the relatedness of stimuli in equivalence classes members was an inverse function of number of nodes separating them holding relational type constant, either transitive or equivalence. Also transitive relations were preferred to equivalence relations holding nodal spread constant. This experiment showed that the relatedness was influenced by the combination of nodal number and relational type. After college students learned two 7-node 9-member equivalence classes, preference tests pitted a 1-node equivalence relation against 1-, 2-, 3-, 4-, and 5-node transitive relations. A 1-node transitive relation were always preferred to a 1-node equivalence relation. Preference for the 1-node equivalence relation was a direct function of nodes in a competing transitive relation, ending in complete preference for the equivalence relation, which showed a strong post class formation graded nodal distance effect. This outcome documented attention to nodal number and relational type in combination. In addition, the equivalence and transitive relations were equally preferred at some intermediate number of nodes in the transitive relation, indicated by the perpendicular in each figure for each class. This value defined the number of nodes needed to weaken a transitive relation to equal the strength of a 1-node equivalence relation. The intactness of the equivalence classes was not influenced by the intervening preference test. Thus, equal and differential relatedness coexisted, with the former and latter being expressed during class formation tests and within class preference tests, respectively.|
|Generalization of an ABA Intervention: A Systematic Generalized Equivalence Class Strategy|
|LANNY FIELDS (Queens College, City University of New York), Kenneth F. Reeve (Caldwell University)|
|Abstract: ): Efficacy of an ABA intervention is indexed by reliable acquisition of some target behavior in the presence of a relevant Sd, and reliable emergence of the target behavior in the presence of novel stimuli that share relevant features with the training exemplars. While acquisition of skills is widely documented in the research literature, generalization is not always reliable. The latter shortcoming is often attributed to training with an insufficient number of exemplars. That explanation, however, does not inform an identification of the factors that could ensure generalization. We will provide a framework addressing this matter that is based on studies of generalized equivalence classes conducted in basic research settings. When clinicians select stimuli to teach a skill, it is often assumed that the stimuli are closely related to each other; i.e., that they are assumed to be members of the same perceptual, open-ended, resemblance-based, fuzzy, or polymorphous class. If stimuli are, in fact, members of a class, and one class member acquires an additional function by training, all class members will reliably evoke that response without additional training. Since ABA interventions typically do not pre-experimentally assess for class membership of the stimuli used to teach and test for skills, unreliable generalization should not be surprising. Alternatively, confirming that stimuli to be used for teaching already function as a class independent of the skill to be taught should greatly enhance generalization. Protocols for the establishment of stimulus classes will be considered as vehicles to enhance the generalization to “novel exemplars” after the training of a target response. This approach is an example of translational research|