|
| Resurgence and Derived Relational Responding |
| Thursday, November 29, 2001 |
| 2:00 PM–2:50 PM |
| Palladian Refectory Hall |
| Area: EAB |
| Chair: Deirdre Beebe Fitzgerald (Eastern Connecticut State University) |
| |
| Fluency and the Facilitation of Derived Relational Responding |
| Domain: Applied Behavior Analysis |
| DEIRDRE BEEBE FITZGERALD (Eastern Connecticut State University) |
| |
| Abstract: The effects of fluency in trained relations on subsequent derived relational behavior was examined to test the claim that fluency in the component skills of a repertoire would produce untrained increases in composite skills. First, the fluency of conceptual experts in stimulus equivalence and matching-to-sample methods was assessed in order to produce a fluency criterion for later conditions. The effect of fluency in the acquisition of conditional relations, symmetric relations, and equivalence relations on derived relational responding in subsequent stimulus classes was examined. Performance of participants who worked to a fluency criterion was contrasted with another overlearning group whose performance was yoked to the fluency group in number of trials, but not rate or accuracy criteria. A final control condition worked only to mastery. A second group of control participants who did not receive training for the derived relations of symmetry and equivalence was included for each of the three learning conditions. Results indicated that training exemplars of the derived relations of symmetry and equivalence did not produce significant increases in the emergence of these relations. Additionally, providing training to a fluency criterion did not produce significant increases the in the emergence of derived relations as compared to either the yoked practice overlearning condition or the mastery only condition. Finally, how readily contextual control could be established given the different learning histories were examined. Contextual control over responding was established in all participants. Derived relations predicted from this training emerged at an equal probability across groups. Implications for both instructional design and descriptions of effective behavior in novel settings are addressed. |
| |
| Resurgence and Recency of Practice |
| Domain: Applied Behavior Analysis |
| FRANCIS MECHNER (The Mechner Foundation), Laurilyn Dianne Jones (The Mechner Foundation) |
| |
| Abstract: Resurgence is defined as the reappearance of antiquated behavior patterns (those observed earlier in a subject's learning history). How recently a skill has been practiced has been shown to be an important variable in determining which behavior patterns will resurge. In several experiments, subjects typed non-word sequences of letters on a computer, each containing both criterial (required) and non-criterial (optional) keystrokes. Several "history" sessions were followed by a final "test" session in which subjects chose which patterns of letters to type. Three groups consisting of three patterns each were learned, separated in some experiments by elapsed time (days without a session) and in others by sessions spent learning unrelated but similar patterns. The objective was to determine the effect of recency of practice on resurgence during the final session. In both experiments, the largest resurgence occurred for patterns from the first and last groups, with almost no resurgence of the middle group of patterns, a finding that is in line with research from the serial learning literature. In further experiments, prefacing the presentation of the first group of patterns with a single session spent learning unrelated patterns was sufficient to wipe out almost completely any preference for the first group. |
| |
|
| |