SIGNALLED REINFORCEMENT AND MULTIPLE SCHEDULES¹

Abstract

The responses of four pigeons were first reinforced in the presence of two different wavelengths (green and red) on a two-ply multiple schedule with identical variable-interval 3-min schedules of reinforcement associated with each component. While the constantcomponent reinforcement schedule remained unchanged during the experiment, the schedule associated with the variable component was changed to (1) signalled variable time, (2) unsignalled variable time, or (3) signalled variable interval. The probability with which the availability of the reinforcer was signalled in the variable-interval schedules was either 0.5 or 1.0. Positive contrast occurred in both signalled variable-interval and variable-time schedules, but only when the availability of all the variable-component reinforcers was signalled. Signalling the availability of only 50% of the reinforcers in signalled variable-interval schedules resulted in negative induction. The present data suggest that positive behavioral contrast resulting from signalled reinforcer availability is due to the presence of an extinction-correlated stimulus. sponse rate of the two components covary in the same direction, the interaction is labelled "positive induction" (rate increases in both components) or "negative induction" (rate decreases in both components). Reynolds (1961a) proposed that the occurrence of positive and negative contrast is a function of the relative reinforcement frequency of the two components. That is, the response rate in the constant component varies inversely with the reinforcement frequency in the variable component. Decreasing the reinforcement rate in the variable component results in positive contrast, while increasing the variable-component reinforcement rate results in negative contrast. Indeed, subsequent research has shown that the relative frequency of reinforcement in the multiple components is a powerful predictor of the magnitude of contrast effect (Bloomfield, 1967;Nevin, 1968; Reynolds, 1961b;Reynolds and Limpo, 1968).An alternative position has been offered by Terrace (1972). He proposed that the necessary and sufficient condition for occurrence of positive contrast is a suppression in the response rate of the variable component, which results as an emotional reaction to the discrimination situation. In this formulation, decreasing reinforcement density is understood to be only 199 NUMBER 2 (SEPTEMBER) 1976, 26,[199][200][201][202][203][204][205][206]

Temporal comparator rules and responding in multiple schedules

“…The main finding of the present study is also surprising in that it contrasts previous research on signaled nondifferential procedures (e.g., Cooper et aI., 1990;Durlach, 1983;Farley, 1980;Marcucella, 1976). In those studies, the overall rate of reinforcement was held constant but cues were added to signal reinforcers that occurred outside the target cue.…”

“…These findings have come from both the autoshaping and multiple-schedule paradigms (Brandon, 1984;Cooperet al, 1990;Durlach, 1983;Farley, 1980;Goddard & Jenkins, 1987;Marcucella, 1976). Farley (1980, Experiment 1; see also Brandon, 1984) examined maintained keypecking in pigeons trained on a three-cue multiple schedule.…”

“…Marcucella (1976) exposed pigeons to a two-component multiple schedule in which both components (e.g., red and green keylights) signaled equal variable-interval (VI) schedules. One component was then changed to a signaled VI schedule, in which a cue was added (a triangle was superimposed on the red keylight), which signaled the availability of reinforcers in that component.…”

“…As in Marcucella's (1976) study, the addition of a cue signaling the added reinforcers created an explicit cue for nonreinforcement (see also Goddard & Jenkins, 1987).…”

Another Look at Contrast in Multiple Schedules

“…The result of that procedure is that responding in the signaled component generally occurs only after the signal has been presented, but the obtained frequency of reinforcement is unaffected. Since contrast has been reported in several studies that have used this procedure (Brownstein & Hughes, 1970;Marcucella, 1976;Wilkie, 1973Wilkie, , 1977, the apparent implication is that response suppression is sufficient to produce contrast, at least under some circumstances.…”