Pieter F. M. Kop

Operant-Pavlovian interactions: The effect of the number of stimuli with free-food depends on the operant baseline

Four groups of three pigeons were trained to respond under operant schedules which varied with respect to response rate requirements and reinforcement frequency. The number of short duration stimuli preceding free food was manipulated in different experimental conditions. Operant behavior was maintained on one key, while stimuli were presented on another key. Stimulus key response rates, latencies to the first response on the stimulus key and the percentage of stimuli responded to did not vary as a function of the number of stimuli presented, but were dependent upon the reinforcement frequency of the operant baseline.

Stimulus- and feeder-directed behavior in a long-box: Effect of fixed versus variable time schedules of food presentation

Two groups of 6 pigeons were exposed to either a fixed-time (FT) or a variable-time (VT) schedule of response-independent food presentation. The interval between two successive food presentations (food-food [FF] interval) was either 20 or 40 sec. The duration of stimulus presentation (stimulus-food [SF] interval) was varied relative to the FF interval. All subjects were exposed to different information ratios (IRs; IR = SF/FF; IR = 1.00, 0.50, 0.15). To study the relative contribution of sign- and goal-tracking behavior to keypecking observed in standard autoshaping procedures, pigeons were autoshaped in a long-box. In the long-box, the stimulus key and the feeder are located 60 cm apart. Stimuli were always presented at one end of the box, and food was presented at the other end. Locomotor behavior and keypecks were recorded. Pigeons engaged in sign-tracking behavior when IR = 0.15, but only when presentation of the food was unpredictable on the basis of other variables (e.g., the passage of time since the last food presentation, as in FT schedules). In the case of FT schedules, subjects engaged in feeder-directed activities. No effects of varying the FF interval were found. Keypecking was observed only when the SF interval was short (IR = 0.15) and food was presented on a VT schedule.

Concurrent schedules: Effects of reinforcement rate and changeover delay on time allocation in a three-alternative procedure

Pigeons were exposed to a continuous choice procedure where three alternatives alternated in a fixed, recycling order (ABCABC, etc.). Responses were reinforced according to independent variable-interval schedules. For three birds, the reinforcement rate for responses on alternative C was varied. For three other birds, the duration of the changeover delay after the changeover to C was varied. For both groups, the reinforcement rates and changeover delay durations associated with A and B were constant throughout the experiment. The time proportion at A relative to B increased as a function of the reinforcement rate for responses on C and decreased as a function of the duration of the changeover delay during C. The results show that the proportion of time spent at a variable-interval alternative of a continuous choice procedure is not completely determined by the reinforcement rates provided by the alternatives. The results support the assumption that time allocation is governed by delayed reinforcement of changeover behaviour.

Concurrent Schedules: Effects of Blackout during the Changeover Delay*

Four homing pigeons were exposed to concurrent variable-interval 40-s variable-interval 120-s schedules of reinforcement with changeover delays of five durations (0, 1, 2, 4, and 8 s). The changeover delays of non-zero durations were presented with and without blackout of the experimental chamber in different conditions. Blackout eliminated or strongly reduced responding during the changeover delay. It decreased the run length of responding after the changeover delay in the variable-interval 120-s component, but it had no systematic effects on the run length after the changeover delay in the variable-interval 40-s component. In both components, the run lengths after the changeover delay increased with the duration of the delay. Blackout increased deviations from response matching, computed from measures of responding after the changeover delay. The deviations increased with the duration of the changeover delay. The reults suggest that the finding of overmatching with procedures other than the changeover delay does not depend exclusively on the absence of responding in the interval that elapses between initiation of the change to an alternative and the first response eligible for reinforcement.

Dissociation of instrumental and Pavlovian contingencies in a discriminated instrumental procedure

Four pigeons were exposed to an experimental procedure in which the presentation of a stimulus at one intelligence panel in a longbox signaled the opportunity to obtain response-dependent food contingent upon withdrawal from the stimulus to the other end of the box. Four other pigeons in addition received unsignaled, response-independent food during the intertrial interval. Both groups of subjects learned to withdraw from the stimulus during stimulus presentations, but the subjects who received response-independent food during the intertrial interval learned to do so faster than the subjects in the other group. Subjects who received response-independent food in the absence of the stimulus hardly ever contacted the stimulus upon its presentation, whereas subjects in the other group (food only presented upon withdrawal from the stimulus) contacted the stimulus frequently. Withdrawal behavior from the stimulus decreased when it was no longer reinforced, while random presentations of the stimulus and response-independent food eliminated stimulus contact behavior.

Operant-Pavlovian interactions: Ratio-schedules and the effects of the duration and location of a stimulus preceding response-independent food

Two groups of four pigeons were exposed to either a fixed-ratio or a random-ratio schedule of reinforcement. Presentation of stimuli which preceded response-independent food delivery was then superimposed upon this operant baseline. The duration and the location of stimulus presentation were varied in different experimental conditions. Random-ratio schedules maintained higher response rates than fixed-ratio schedules. Response rates during presentation of the stimulus on the operant key exceeded response rates in the absence of stimulus presentation for the fixed-ratio subjects, but not for the random-ratio subjects. Similar observations were made when stimulus presentation occurred on a seperate stimulus key. During stimulus presentation on the stimulus key, response rates on the operant key were considerably higher for the random-ratio subjects than for the fixed-ratio subjects, especially during longer presentation of the stimulus (30 sec).