Stimulus functions in some chained fixed-ratio schedules of reinforcement.
Load most responses into the final link of a chain to reduce early pausing, but expect longer pauses before that final step.
01Research in Context
What this study did
Two pigeons worked on chained fixed-ratio schedules. Each chain had two links. The first link always needed 2 pecks. The second link needed 2, 10, 50, or 98 pecks.
The birds earned grain only after finishing both links. Researchers timed how long the birds paused before starting each link.
What they found
When the second link needed 98 pecks, birds paused less before the first 2-peck link. But they paused more before the big 98-peck link.
The 2-98 chain gave the shortest pause at the start. Yet it created the longest pause at the end.
How this fits with other research
FARMEMOORHEARSKELLEHER et al. (1964) saw the same pattern years earlier. They used variable-interval schedules and found birds sped up as the final reward got closer.
Byrd (1972) tested fixed-interval chains and showed choice depends on absolute time, not just ratios. This matches our finding that pause length tracks the actual number of responses needed.
Cohen (1975) proved birds can tell ratio sizes apart. This supports why pause times change when the second link grows from 2 to 98 pecks.
Why it matters
When you build token boards or sticker charts, put most work near the final prize. This cuts early pausing. But know the last step may take longer. Test both arrangements with your learner to find the sweet spot.
Get CEUs on This Topic — Free
The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.
Try a 2-response first link and 10-response second link in your token board, then time how long the learner takes to start each step.
02At a glance
03Original abstract
For two pigeons, behavior was compared in equivalent two-link chained and tandem schedules in which 100 responses were required for food reinforcement but the responses required in the two links of the chained schedule were varied over the values 2-98, 50-50, and 98-2. Initial pausing, response rate in each link, and the total time to complete each ratio were recorded. In chained components, initial-link pausing (for both birds) and total time to complete the ratio (for one bird) were generally shorter when the response requirement was 2-98 than when it was 50-50 or 98-2; for terminal-link pausing this relationship was reversed for both birds. There were also systematic changes in behavior in tandem components, and in the relationship between behavior in chained and tandem components. The results are discussed in terms of conditioned reinforcement.
Journal of the experimental analysis of behavior, 1971 · doi:10.1901/jeab.1971.15-311