MacDonall, J. S. (2000).
Synthesizing concurrent interval performances.
Journal of the Experimental Analysis of Behavior,
74, 189-206.
Concurrent schedules may be viewed as consisting of two pairs of
stay and switch schedules, each pair associated with one of the
alternatives. A stay schedule arranges reinforcers for staying
and responding at one alternative, whereas the associated switch
schedule arranges reinforcers for switching to the other
alternative. In standard concurrent schedules, the stay schedule
at each alternative is equivalent to the switch schedule at the
other alternative. MacDonall (1999) exposed rats to one pair of
stay and switch variable-ratio schedules and varied the response
requirements across conditions. Combining results from symmetric
pairs produced composite performances that were described by the
generalized matching law. This outcome was noteworthy because the
data were obtained from performances at two alternatives with
contingencies that were functionally unrelated to each other.
This result suggests that concurrent performances may consist of
two unrelated performances that alternate as behavior moves
between alternatives. The purpose of the present experiment was
to extend those results to interval schedules. Rats were exposed
to pairs of random-interval schedules, and across conditions
their mean intervals were varied. When data from appropriately
paired conditions were combined, the composite performances were
consistent with the generalized matching law. In addition, the
results supported two models of concurrent performances that were
based on local variables at an alternative (behavior, and stay
and switch reinforcers): a modified version of the contingency
discrimination model (Davison & Jenkins, 1985) and the local
model (MacDonall, 1999).
Key words: generalized matching law, contingency discrimination
model, local model, melioration, random-interval schedule, lever
press, rats