Temporal factors influencing the acquisition and maintenance of an autoshaped keypeck

Abstract

Food-deprived pigeons were given brief meals of grain following the presentation of a light on a response key. Pecking the key had no consequence. Virtually all of the pigeons pecked the ligated key. The number of trials prior to the first peck varied inversely with the value of the mean interval between light onsets. Trials to criterion was a negative power function of the value of the intertrial interval. The addition of a second stimulus, never followed by food, retarded the acquisition of toe keypeck, particularly at short intertrial intervals. During steady state performance, the value of two measures of response strength, rate and probability of responding, increased as a function of the duration of the intertrial interval.

Determinants of pigeons’ choice between certain and probabilistic outcomes

“…At the same time, the duration of the ITI was not completely without effect on behavior, because rate of responding to the terminal-link stimuli associated with food on both sides was higher when the ITI was long than it was when the ITI was short. This result is consistent with findings from classical conditioning studies showing that rates of elicited responding to a CS for food increase as a function of ITI (e.g., Terrace, Gibbon, Farrell, & Baldock, 1975). The finding that the ITI affected absolute rates of responding to both food terminal links, but failed to alter either the relative rates to the 50% versus the 100% food terminal links or the choice proportions for the 50% alternative, suggests that the ITI may have affected equally the value of both the 100% terminal-link stimulus and the 50% Sϩ, hence leaving their relative values unchanged.…”

“…That is, the subjects responded at higher rates when opportunities to choose were less frequent. This enhancement of responding to stimuli associated with food when ITI is lengthened is consistent with trial-spacing effects on autoshaped response rates (e.g., Terrace et al, 1975). The effect of ITI duration on absolute response rates during food-associated stimuli, in the context of no effect on either preference between the alternatives or relative terminal-link response rates, suggests that lengthening the ITI may have equally enhanced the value of all periods of time associated with food.…”

Partial reinforcement in autoshaping with pigeons

“…For example, effects of massed vs. distributed training on autoshaped responding (Gibbon, Baldock, Locurto, Gold, & Terrace, 1977;Terrace, Gibbon, Farrell, & Baldock, 1975) have been found to parallel the traditional finding of more effective training with spaced trials. The present experiment is addressed to another traditional variable in classical conditioning, partial reinforcement.…”

One-trial appetitive conditioning in the sexual behavior system

“…More extensive training is also required to condition pigeons to peck at visual signals for food. Terrace, Gibbon, Farrell, and Baldock (1975), for example, reported that, even under optimal procedural parameters, their fastest learners required about 10 trials to acquire autoshaped keypecking (see also Locurto, Travers, Terrace, & Gibbon, 1980).…”