|
| Transitive Repertories, Linked Perceptual Classes and Generalized Equivalence Classes |
| Tuesday, June 1, 2004 |
| 10:30 AM–11:50 AM |
| Fairfax A |
| Area: EAB; Domain: Applied Research |
| Chair: Antonios Varelas (Queens College/SUNY) |
| Abstract: . |
| |
| Transitivity-indicative Performances During and After Two Transitivity Induction Procedures |
| ANTONIOS VARELAS (Queens College/SUNY), Lanny Fields (Queens College/SUNY) |
| Abstract: Twenty-four college students failed to demonstrate transitivity-indicative performances in an INITIAL TEST that included six stimulus sets drawn from three visual domains all of which were presented in a randomized order in a single block of trials. Thereafter, half of the participants were given SERIALIZED RE-EXPOSURE to the same stimuli that were presented in the initial test, but in a series of test blocks, and in the absence of feedback. The other participants were given TRANSITIVITY TRAINING with 20 new stimulus sets. When given SERIALIZED RE-EXPOSURE resulted, 83% of the participants showed a substantial increase in transitivity-indicative performances WHICH was maintained when presented with a replication of the initial test. During TRANSITIVITY TRAINING, although only 35% of the participants showed the emergence of transitivity-indicative performances, 83% of all subjects then showed transitivity-indicative performances during the final test final test which was a replication of the initial test. The bases for such behaviorally silent learning in the absence of differential reinforcement will be discussed. |
| |
| Relational Strength Among Stimuli in Linked Perceptual Classes: A Chronometric Analysis |
| KIMBERLY SHAMOUN (Queens College/CUNY), Danielle Tittelbach (Queens College/CUNY), Lanny Fields (Queens College/CUNY) |
| Abstract: A linked perceptual class consists of two disparate perceptual classes, the members of which come to occasioned the mutual selection of each other. This occurs are the establishment of at least one conditional discrimination among single members of each of the classes. First, we identified three key stimuli in each of four distinct perceptual class. One endpoint of a class, called the anchor, was the clearest member of that class. The other endpoint of the class, called the boundary, was the most perceptually distant from the anchor that still functioned as a class member. The third, called the midpoint, was the stimulus perceptually equidistant from the anchor and boundary stimuli. The anchors, midpoints, and boundaries for were designated for four classes: A1a, A1m, and A1b for class A1`, A2a, A1m, and A2b for class A2`, B1a, A1m, and B1b for class B1` along with B1a, A1m, and B2b for class B2`. The A` and B` classes were linked in one of six training conditions. Aa-Ba and Ab-Bb, Aa-Bb and Ab-Ba, Aa-Ba only, Ab-Bb only, Aa-Bb only, and Ab-Ba only. The emergence of a linked perceptual class was documented when all novel combinations of anchor, midpoint and boundary stimuli from two linked classes occasioned the selection of each other. With five of the six training conditions, the reaction times were invariant across all of test combinations used to demonstrate the formation of the linked perceptual classes. After Aa-Ba training, however, reaction times varied systematically with the anchor, midpoint and boundary values of the stimuli used in the tests. These data, then, show that all of the stimuli in a linked perceptual class are not functionally interchangeable. That interchangeability depends on the way in which the two perceptual classes are linked by training. |
| |
| The Expansion of Linked Perceptual Classes to a Generalized Equivalence Class |
| PATRICIA A. MOSS-LOURENCO (Queens College/CUNY), Adrienne Fitzer (Queens College/CUNY), Lanny Fields (Queens College/CUNY) |
| Abstract: A fully elaborated generalized equivalence class consists of at least three perceptually disparate stimuli that function as members of an equivalence class, A-B-C, along with the dimensional variants of each member of the basal equivalence class, A`, B`, and C`, all of which occasion the mutual selection of each other, A`-B`, B`-A`, A`-C`-C`-A`, B`-C` and C`-B`, in the absence of direct training. To date, fully elaborated generalized equivalence classes have not be established in laboratory conditions. In the current experiment, we approximated the formation of fully elaborated generalized equivalence classes in the following manner. First we determined the endpoints of four perceptual classes A1`, A2`, B1`, and B2`. The endpoints of which were called the anchors (Xa) and boundaries (Xb). After training conditional discriminations between the anchor stimuli of an A and B class and the boundary stimuli of the same A and B classes, all of the stimuli in the linked classes occasioned the mutual selection of each other thereby documenting the emergence of two linked perceptual classes, A1`-B1` and A2`-B2`. Then, C1 and C2 stimuli, which were perceptually distinct from each other and from the stimuli in the A1`, A2`, B1` and B2` classes, were linked to single members of one linked perceptual class through B1a-C1 and B2a-C2 training. Finally, new tests showed that all member of the A1` and B1 classes occasioned selection of the C1 stimulus and vice versa. Likewise for the class 2 stimuli. These results demonstrated the formation of a partially elaborated generalized equivalence class, A`-B`-C, which approximate fully elaborated generalized equivalence classes. The formation of these classes was interfered with by delaying the tests for expansion of class size, and was enhanced by documenting the intactness of the linked perceptual class before measuring expansion of class size. |
|
| |