|
Looking Into Details: Delayed Reinforcement, Behavioral Variability, Continuous Repertoires and Stimulus Control |
Sunday, May 26, 2013 |
10:00 AM–11:20 AM |
101 H (Convention Center) |
Area: EAB; Domain: Basic Research |
Chair: Iver H. Iversen (University of North Florida) |
Abstract: Delayed reinforcement, lag-N contingencies, and multiple exemplar training have all been applied in many behavior-analytic studies. The papers presented in this symposium are concerned with looking more into the details of how behavior is affected under variations of these well-known procedures. |
|
Analysis of Contingencies of Delayed Reinforcement |
IVER H. IVERSEN (University of North Florida) |
Abstract: Acquisition of operant behavior with delayed reinforcement has been established with delays up to 30 s for several species and types of reinforcement. However, how subjects respond to changes in contingencies of reinforcement after behavior has been acquired with delayed reinforcement has not been a topic for research. Rats were used as subjects, and a nose-poking response (without response feedback) produced a single food pellet after an 8-s resetting delay. After acquisition, some rats were exposed to a Fixed-Duration (FD) schedule with a gradually increasing response duration requirement up to 1 s while other rats were exposed to Fixed Ratio (FR) up to 5. Then, the delay was removed while the schedule was retained. Next, the 8-s delay was reinserted. After acquisition with delayed reinforcement, behavior was maintained under FD or FR but in a very inefficient manner with considerable response variability. When the delay was removed behavior improved in efficiency and variability was reduced but not as much as for comparison rats who acquired the same response under FD 1 s or FR 5 without delay during acquisition. Contingencies of delayed reinforcement are complex and induce considerable response variability that does not easily recover when delays are removed. The paper will present various methodological issues related to studying the effects of changing contingencies of delayed reinforcement. |
|
Behavioral Variability as an Operant Dimension? |
LINE FLATEBØ WIDMARK (Oslo and Akershus University College), Per Holth (Oslo and Akershus University College) |
Abstract: Variation is the raw material on which selection operates. Behavior variability is a prerequisite for learning to occur, as operants are selected from variation by their consequences. Some well-known experiments have suggested that variability can be directly reinforced and can function as an operant class of its own, and that organisms can engage in the strategy of a quasi-random generator. One aim of the present study was to investigate the occurrences of increased and decreased response variability under different lag contingencies in rats. A second purpose was to study the immediate effect of reinforcement of new responses. After a baseline of reinforcement of four different response topographies according to a lag 3 contingency, we introduced a fifth operandum and changed to a lag 4 contingency. The results indicate that reinforcement typically produces stereotypic responding, while response variability increases during extinction. When reinforcement was contingent on a novel response, immediate subsequent responding was not novel or varied, but repetition of the response topography that produced reinforcers. |
|
Establishment of a Continuous Repertoire in Rats |
MORTEN BERGER (Oslo and Akershus University College), Per Holth (Oslo and Akershus University College) |
Abstract: In a continuous repertoire, a response dimension varies as a function of changes along some stimulus dimension. Such repertoires include a spread of effect from trained to untrained exemplars. However, the empirical evidence for such repertoires is sparse. The few experiments that have demonstrated continuous repertoires have varied with respect to procedures, findings, and conclusions. The present study showed the establishment of a continuous repertoire in four Wistar rats. The stimulus dimension consisted of seven cue lights arranged in a horizontal row, and the response dimension was given by a horizontal line of seven levers, localized just below the row of lights. The results showed that intermediate response values emerged in the presence of intermediate stimulus values, and that responses outside of the trained range emerged in the presence of corresponding stimuli outside the ranges of stimuli during training. |
|
Teaching Rats to Play the Keyboard: Stimulus Control Concerns |
PER HOLTH (Oslo and Akershus University College) |
Abstract: Four rats participated in an experiment on continuous repertoires. The apparatus used in the experiment was a standard operant experimental chamber supplied with seven adjacent levers that could serve as a keyboard, and sounds of different pitches were programmed as immediate consequences of lever presses. After participating in the experiment on a continuous repertoire, random sequences of lights were programmed, and corresponding sequences of responses (light following) were intermittently reinforced. Different tones were programmed as consequences of pressing the seven levers as a keyboard. Finally, a sequence of lights was programmed to indicate the sequence of levers to be pressed to produce Pink Floyd’s “Another Brick in the Wall” (“We don’t need no education”), and then “White Christmas”. Video demonstrations show the rats’ final performances. |
|
|