ps

Senior Theses: 1997-1998

The Development of Functional Response Units
Cynthia Z. Chadwick
Faculty Advisor: Alliston K. Reid, Ph.D.

These data have been published in the Journal of the Experimental Analysis of Behavior. 

In order to study the processes responsible for the development of functional response units, rats were exposed to a probabilistic free-operant schedule. Changeovers from one lever to another were reinforced with greater probability than was preservation on the same lever. An imaginary coin was flipped after every even lever press to determine whether or not to reinforce the undemarcated response sequence, but no other accompanying changes occurred in the apparatus. Subjects first came under control by the differential reinforcement contingency, followed by stable transition probabilities across levers. Analysis of interresponse times (IRTs) demonstrated that, with extended training subjects organized their responding into two-response sequences framed by the tosses of the imaginary coin. When a tone accompanied each coin toss in a second condition, differences in the IRTs following the first and second response became even greater. Sequence production approximately matched relative reinforcement rates. These findings suggest that models of learning must include a mechanism for changing the response unit as a function of reinforcement history.

The Identification of Behavioral Units in Learned Response Sequences in Rats and their Dependence on Trial Cues
Stephen Gray
Faculty Advisor: Alliston K. Reid, Ph.D.
These data have been published in book chapter by MIT Press.

The present study was interested in the mechanisms that influence the production of new sequences and what effects, if any, do end-of-trial cues have upon those mechanisms. To this end, six rats were trained to complete three-response sequences to obtain food. Each subject was trained on six different sequences. A tone accompanied every food presentation. During the test phase of the experiment, probe trials were inserted in which there was no food presented and the tone was presented at random. The sequence produced after the probe trial was what this study was most interested in. Results found suggest that the dominant response of a sequence plays a pivotal role in the production of new sequences. Also, a metric for the prediction of sequences created was examined. The results support this metric that is found in table 5 (see Table 5).

Response Strength in a Multiple Schedule of Reinforcement
Jill Kelley, Michelle Rayborn, & Stephen Renna
Faculty Advisor: Alliston K. Reid, Ph.D.
These data have been submitted for publication to the Journal of the Experimental Analysis of Behavior. 

Four pigeons were trained to peck two keys. Two birds were required to peck the red key followed by the white key in order to be reinforced. The remaining two birds were required to peck the white key followed by the red key in order to be reinforced. The purpose of the experiment was to determine which response was more closely associated with the reinforcer and to measure the resistance to change of each response when new variables were added. The multiple schedule consisted of two components. In Component A, reinforcement was received only if the pigeons' responses matched the previously learned sequence. Component B was a variable interval schedule in which only the center light was illuminated. The components were separated by a blackout phase, which was later used to test resistance to change. This experiment is currently underway, and all pigeons have reached baseline on their particular schedules, allowing them to be switched to a multiple schedule.

The Effect of Tactile Properties on Tremor in Lever Presses by Human Subjects
Mary J. Carmichael, C. Elaine Fincher, & Aubrey K. Senn
Faculty Advisor: Don Scott, Ph.D.
 

Modified tactile surfaces increase hand steadiness and accuracy when a lever is pressed and held for two seconds. Four surfaces were cast from liquid casting resin and contained protrusions 5 or 8 mm long that were formed from wooden golf tees, and terminated in 1 mm or 5 mm cross-sectional area tips. Lever pressure was sampled at 64 Hz and presses were Fourier analyzed for physiologic tremor. Results showed a significant reduction in the 3-5 Hz (Cerebellar) tremor frequency due to the Long (8 mm)/Sharp (1 mm area) surface. The greatest tremor was due to the Short (5 mm)/Dull (5 mm area) surface.

Response Strength in Response Sequences
Matt Cunningham, Michelle Rayborn, & Patricia Shull
Faculty Advisor: Alliston K. Reid, Ph.D.
These data have been submitted for publication to the Journal of the Experimental Analysis of Behavior. 

The mechanisms responsible for response strength within response sequences were studied in this experiment. The study investigated positional differences in response strength and the contingencies forming these strengths. Four pigeons were trained to peck colored keys in order to receive reinforcement. The pigeons were placed on a multiple schedule which included VI 30-second white and VI 30-second red schedules which were separated by blackout and sequence components. The response rates of the two different VI 30-second schedules were compared to show differences in response rates created by operant contingencies. There were no major differences between the two response rates and the two seemed to correlate, showing the absence of operant contingencies. The experiment is still ongoing. The Pavlovian contingencies will be tested by administration of free food and testing for resistance to change.

The Discriminative Stimulus in Learned Response Sequences
Corey Weaver, Sarah Hixon, & Wells Shepard
Faculty Advisor: Alliston K. Reid, Ph.D.
 

Six white rats were trained on a three-response sequence for reinforcement. A tone was placed between the first and second lever press during fifty percent of the trials in Condition 1 to acquire discriminative stimulus properties. The tone indicated to the subject their next response in the sequence. In Condition 2, the tone was moved between the second and third position for about fifteen trials. A paired sample t-test indicated a significant degree of stimulus control by the discriminative stimulus in determining the next response made by the subject.