Preference in concurrent variable-interval fixed-ratio schedules.
A two-part ratio model predicts choice better than the simple matching law when reinforcers come at different speeds.
01Research in Context
What this study did
Pigeons pecked two keys. One key paid off on a variable-interval schedule. The other paid off on a fixed-ratio schedule.
The researcher tracked which key the birds chose. He then compared two math models. One was the basic matching law. The other was a new dual-sensitivity model.
What they found
The dual-sensitivity model fit the data better. It predicted how the birds split their time between the two keys. The old matching law missed the mark.
How this fits with other research
Lea et al. (1977) showed that pigeons work less for plain pellets when tasty sucrose is also available. Durand (1982) adds a math tool that captures that drop in work.
Cullinan et al. (2001) later used signaled delays and still saw strong preference. Their data also follow the dual-sensitivity pattern, even though the task changed.
Rojahn et al. (1994) tested ratio versus difference rules. They found ratio rules win. Durand (1982) uses ratio math too, so the two papers agree that animals compare rates, not raw numbers.
Why it matters
If your client has two ways to earn reinforcement, the basic matching law may underestimate which option they pick. Try plotting the dual-sensitivity model. It can show why a child stays on a tablet longer than your data sheet predicts.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Graph response splits during free-operant preference tests; if the basic matching line drifts, plug the same data into a dual-sensitivity formula and adjust the menu.
02At a glance
03Original abstract
Five pigeons were trained on concurrent variable-interval fixed-ratio schedules in three experiments. Experiment 1 used two variable-interval schedules and one fixed-ratio schedule, and the ratio requirement was varied. Using the generalized matching law, sensitivity to reinforcement was close to 1.0, but performance was biased toward the variable-interval schedule with the lower reinforcement rate. In Experiment 2, which used one variable-interval and one fixed-ratio schedule, the interval schedule was varied. All birds showed sensitivities to reinforcement of less than 1.0 and of less than the values obtained in Experiment 1. The performance was also biased toward the fixed-ratio schedule. Because the generalized matching law could not account for the differences in the data from Experiments 1 and 2, an extension of this law was suggested and successfully tested in Experiment 3. The proposed dual-sensitivity model was also shown to clarify some previously reported results.
Journal of the experimental analysis of behavior, 1982 · doi:10.1901/jeab.1982.37-81