SECOND‐ORDER SCHEDULES OF TOKEN REINFORCEMENT: COMPARISONS OF PERFORMANCE UNDER FIXED‐RATIO AND VARIABLE‐RATIO EXCHANGE SCHEDULES¹

Abstract

Rats' lever pressing produced tokens according to a 20-response fixed-ratio schedule. Sequences of token schedules were reinforced under a second-order schedule by presentation of periods when tokens could be exchanged for food pellets. When the exchange period schedule was a six-response fixed ratio, patterns of completing the component token schedules were bivalued, with relatively long and frequent pauses marking the initiation of each new sequence. Altering the exchange period schedule to a six-response variable ratio resulted in sharp reductions in the frequency and duration of these initial pauses, and increases in overall rates of lever pressing. These results are comparable to those ordinarily obtained under simple fixed-ratio and variable-ratio schedules.Key words: second-order schedules, token reinforcement, variable-ratio schedules, fixedratio schedules, lever pressing, ratsUnder second-order chained and brief-stimulus schedules, behavior that satisfies the requirements of a component schedule is treated as a unitary response that is itself reinforced according to some schedule of reinforcement (cf. Kelleher, 1966). Under second-order token schedules, performance within component schedules of token delivery is reinforced according to the schedule for presenting the opportunity to exchange tokens for (usually) food (Malagodi, Webbe, and Waddell, 1975). Operationally, the delivery of tokens may be likened to brief presentations of exteroceptive stimuli, as in second-order brief-stimulus schedules, or, alternatively, the accumulation of tokens may be likened to successive changes in discriminative stimuli as in second-order chained schedules. Direct experimental comparisons of these three forms of second-order schedules have not been reported, but performance under token schedules most often resembles brief-stimulus rather than chained schedule performance (Malagodi et al., 1975b;Waddell, Leander, Webbe, and Malagodi, 1972).Because the completion of component schedules may be conditionable with respect to second-order schedule dependencies, second-order schedules have become important procedures for examining the generality of the effects of variables that contribute to the rates and patterns characteristic of schedule-controlled behavior in general (e.g., Davison, 1969;Marr, 1971; see Gollub, 1977 for review). The use of second-order schedules within such an analytical context provides a means for extending the generality of schedule processes and the unitary effects of schedule variables from simple to complex behavioral situations (cf. Morse, 1966). For example, under simple fixedratio (FR) and fixed-interval (FI) schedules, the positive, direct relationship between postreinforcer pause duration and schedule parameter value (Felton and Lyon, 1966;Schneider, 1969) is replicated under second-order FR and Fl schedules (Malagodi; 1967b, c;Waddell et al., 1972). Under simple fixed schedules at parameter values that generate marked postreinforcer pauses, changing the schedule dependency from fixed to vari...

The several roles of stimuli in token reinforcement

“…The temporal organization of behavior across the token-reinforcement schedules in the first experiment is similar to that seen in other token-reinforcement schedules with FR tokenproduction and exchange schedules (Kelleher, 1958;Webbe & Malagodi, 1978;Foster et al, 2001;Bullock & Hackenberg, 2006) and in extended-chained schedules with FR components (Jwaideh, 1973). The present experiment sought to more systematically compare tokenreinforcement and extended-chained schedules with respect to both overall response rates and temporal organization of behavior.…”

Generalized conditioned reinforcement with pigeons in a token economy

“…While extending the use of token systems to new domains, the present study also connects to prior work on extended sequence schedules, including but not limited to token reinforcement schedules. As a type of extended chained schedule, token reinforcement schedules produce temporally organized patterns of behavior, with response rates increasing across token‐production segments leading to the terminal reinforcer (Bullock & Hackenberg, ; Foster et al, ; Waddell, Leander, Webbe, & Malagodi, ; Webbe & Malagodi, ). The differential latencies shown in Figure are consistent with this general finding: Latencies at the outset of token‐production components were consistently longer than those at the outset of exchange‐production components.…”

Token reinforcement: Translational research and application

“…For example, an early study by Staats, Staats, Schutz, and Wolf () showed that tokens were effective in generating and maintaining behavior in a reading program with children, even with exchange‐production ratios as high as 24 (i.e., tokens could be exchanged after 24 had been earned). Staats et al also reported evidence of break‐run patterns similar to those obtained with FR exchange‐production schedules with chimpanzees (Kelleher, ), rats (Webbe & Malagodi, ), and pigeons (Bullock & Hackenberg, ; Foster et al, ). The similar schedule‐related effects seen across settings and species suggest that token‐reinforced behavior in the laboratory and in applied settings may be governed by similar principles.…”