|
| Equivalence 2 |
| Tuesday, June 1, 2004 |
| 10:00 AM–10:50 AM |
| Gardner |
| Area: EAB |
| Chair: Harold E. Lobo (West Virginia University) |
| |
| Emergence of Equivalence Relations under Requirements of Low Response Latencies |
| Domain: Applied Research |
| GERSON YUKIO TOMANARI (University of Sao Paulo), Murray Sidman (University of Sao Paulo), Adriana Rubio (University of Sao Paulo), William V. Dube (E.K. Shriver Center, University of Massachusetts Medical School) |
| |
| Abstract: In a matching-to-sample task, baseline conditional discriminations between two sets of four stimuli were taught to five human adults who were later tested for the emergence of new conditional discriminations consistent with symmetry and equivalence (combined symmetry and transitivity). During training, the required sample and comparison latencies were gradually decreased, given the maintenance of accuracy (90% correct). The final maximum latencies were 0.4-0.5s to samples and 1.2-1.3s to comparisons. Intertrial intervals were reduced to 0.4s. Under these fast-responding requirements, 72 unreinforced test trials were interspersed among 72 unreinforced baseline trials. Accuracies in the first equivalence and symmetry tests were respectively around 57%-81% (Participant PLRA), 74%-61% (Participant ALE), 76%-80% (Participant CA), and 85%-92% (Participants EMM and SU). The lack of accuracy in the test trials was mostly due to failures to respond within the required latencies, not to incorrect choices. When participants did respond in time, their accuracies were normally higher and around 71%-79% (Participant PLRA), 85%-68% (Participant ALE), 87%-87% (Participant CA), and 96%-97% (Participants EMM and SU). Results show that the emergence of new conditional discriminations and equivalence class formation occurred with requirements for very brief response latencies, a condition in which the possibility of verbal mediation was highly restricted. |
| |
| Studying Problem Solving through Stimulus Equivalence: Solutions as Emergent Blends of Previous Behavior |
| Domain: Applied Research |
| HAROLD E. LOBO (West Virginia University), Philip N. Chase (West Virginia University) |
| |
| Abstract: This paper discusses methodological issues involved in adopting stimulus equivalence procedures for conducting research on problem solving. Problem solving is behavior that produces a change in the environment such that it becomes discriminative for further responding that produces the solution, which was established as a reinforcer for the individual’s behavior. Studies on stimulus equivalence have employed matching-to-sample tasks to train baseline conditional discriminations and to test for reflexivity, symmetry, transitivity, and combined symmetry and transitivity. Changes in the conditional discrimination for each test can be said to arrange a problem, the solution of which is demonstrated by the emergence of untrained conditional relations. Furthermore, the combined symmetry and transitivity tests represent problem solving where independent emergent repertoires are blended as a solution. Stimulus equivalence procedures might be particularly useful for studying problem solving because manipulations of the modality and complexity of the stimuli, the individual’s familiarity with the stimuli, the number of stimuli per class, the training structure, and the total number of classes will render various levels of complexity for the problem-solving arrangement. These procedures also allow repeated presentations of the same problem-solving task. This multiplicity of options may prove fruitful to study the influence of other variables, such as instructions, on problem solving. |
| |
|
| |