The effects of a novel stimulus change on responding in extinction following fixed-ratio training

Bulletin of the Psychonomic Society 1977, Vol. 9 (5), 340-342 The effects of a novel stimulus change on responding in extinction following fixed-ratio training LARRY A. ALFERINK Drake Um'ver8"ity, De8 Moine8, Iowa 50311 and EDWARD K. CROSSMAN Utah State University, Logan, Utah 84322 The present study assessed the effects of a brief stimulus change not paired with food on the pattern of responding in extinction following fixed-ratio (FR) training. Four pigeons were trained to respond on a FR 100 schedule of food reinforcement. Two of these subjects were then exposed to four extinction sessions in which only this novel stimulus change followed completion of each FR 100. The other two pigeons were exposed to six extinction sessions with the unpaired stimulus change following completion of each FR 100 in odd-numbered extinction sessions. In even-numbered sessions, the stimulus change did not occur. A minimum of five retraining sessions occurred between extinction sessions. The results showed that the unpaired stimulus change controlled the location of pausing. Of the total number of pauses recorded during extinction, a greater proportion of these followed FR 100 schedules with stimulus change than without stimulus change. Consideration is given to the properties of the stimulus change responsible for these pause patterns. Experiments using conditioned reinforcement techniques have frequently presented brief stimulus changes which are paired with food _ The pattern of responding generated by these brief stimulus changes, as well as increases in response rate, have been taken as evidence that the stimuli are conditioned reinforcers. For example, Kelleher (1966) studied second-order schedules in which a white key light was presented briefly following the completion of a ftxed-interval (FI) component. Food was delivered following a ftxed ratio (FR) of these stimulus changes. Responding generally was positively accelerated within each interval only if a stimulus change paired with food delivery occurred at the completion of each FI component. However, other studies with second-order schedules (Stubbs, 1971; Stubbs & Cohen, 1972) and with percentage reinforcement (Ferster & Skinner, 1957; Neuringer & Chung, 1967) have found either schedule patterning with unpaired stimuli or no difference in patterning with paired or unpaired stimuli. Similarly, even though a paired stimulus change produces appropriate schedule patterning, the stimulus change may not produce more responses in extinction when compared with extinction sessions without a The authors with to thank Joanne Koenig and Ben Blair for their assistance in training subjects for this experiment. This research was supported by Grant UTP 907-7 from Utah State University . Reprints may be obtained from Larry Alferink, Department of Psychology, Drake University, Des Moines, Iowa 50311. stimulus change (Weissman & Crossman, 1966). Taken together, these results question the pairing hypothesis of conditioned reinforcement. A difficulty with studies which have compared paired and unpaired stimuli is that food and the unpaired stimuli occur in the same session . Thus, a relationship between food delivery and the unpaired stimuli could exist. One way to avoid this difficulty is to present the unpaired stimulus only in those sessions in which food is not delivered, hence the term "novel." The purpose of the present study was to examine the effects of this novel unpaired stimulus on response patterning in extinction following FR training. METHOD Subjects Four experimentally naive White King pigeons served. Their age and sex were undetermined. Water was available in the home cage. Appuatus Two identical three-key pigeon chambers were used. The front panel was covered during experimental sessions by a paper mask such that only the center key, the food hopper located 10.5 cm directly below the key, and two panel lights (G .E. 1820), which provided illumination during experimental sessions. were exposed. Reinforcement consisted of 3.5 sec access (3 sec for Bird K·12) to an illuminated hopper containing Purina Racing Checkers. White noise was present in the room at all times to mask extraneous sounds. The experiment was cuntrolled and data were recorded by electromechanical equipment housed in an adjacent room . 340 EXTINCTION FOLLOWING FR TRAINING Procedure All subjects were maintained at approximately 80% freefeeding weight and shaped to peck a red key. The number of responses required for reinforcement was gradually increased until performance on a FR 100 schedule was established. A relay click followed each response during training sessions except during food delivery. Training remained in effect until the behavior was stable, as determined by visual inspection of cumulative records. During all sessions, subjects were placed in and removed from the test chamber when the key and panel lights were illuminated. The lights remained illuminated at all times except as noted below. Session length during FR 100 session was approximately 100 reinforcers. Sixty-minute extinction sessions followed training on the FR 100 schedule. During these extinction sessions, each completion of 100 responses was followed by a 3.5-sec cycle (3 sec for K-12), the duration of the food cycle during training. During this cycle, responses did not produce a click from the feedback relay. During some extinction sessions, the response key was darkened during the 3.5-sec cycle. For Subjects K-9 and K-12, a total of four extinction sessions were conducted, with the novel stimulus change following each 100 responses. For Subjects K-17 and K-18, six extinction sessions occurred , with the stimulus change programmed only for the first, third, and fifth of the six sessions. During the other sessions, no stimulus change occurred and the key and panel lights remained illuminated at all times. A minimum of five retraining sessions occurred between each of the extinction sessions for all subjects. The conditions during retraining were the same as those in training. RESULTS In extinction, the birds generally ceased responding when the key was darkened after FR 100. Examination of cumulative records showed that , in the stimuluschange extinction sessions, the pause which began at the onset of the stimulus change often continued after the red key light was again illuminated. In extinction sessions without the stimulus change, pauses rarely followed the completion of each 100 responses, but instead occurred at other locations in the ratio. The relationship between the location of pauses and the type of extinction session is shown in Table 1. In this table, the proportion and percentage of pauses after the 3.5-sec cycle to the total number of pauses is shown for each extinction session for each subject. Pauses were measured from cumulative records and a pause was arbitrarily defined as a period of 10 sec or more without a response . The 3.5-sec stimulus-change cycle was exclude (...truncated)