Determinants of choice for pigeons and humans on concurrent-chains schedules of reinforcement.
Pigeons pick the path that shortens the wait; humans pick the path that pays the most overall.
01Research in Context
What this study did
Researchers let pigeons and adult humans pick between two chained schedules. Each chain had a first link followed by a second link that ended in food for birds or money for people.
The team kept the overall rate of payoff the same on both sides. They only changed how long each side's links lasted. Then they watched who picked which side.
What they found
Pigeons followed the delay-reduction rule. They chose the side that cut wait time to payoff the most, even if that side paid less often overall.
Humans did the opposite. They picked the side that gave the higher total payoff per minute. They ignored which side shortened the wait more.
How this fits with other research
Singh et al. (1982) saw the same pigeon pattern earlier. Birds took longer schedules after long search times. The new study shows the rule still holds when humans join the test.
Doughty et al. (2010) later showed pigeons and humans can act alike, but only when the setup uses tokens that swap right away. That finding extends the 1989 result: species differ under standard delays, yet align when you speed up exchange.
Kydd et al. (1982) found pigeons ignore relative immediacy when total delay is fixed. Steege et al. (1989) now show humans still track overall rate, revealing a true species split, not just a procedural quirk.
Why it matters
When you write token boards or chained schedules for clients, remember: people chase the best long-term rate, not the fastest next step. To keep kids engaged, highlight total possible points per session, not just the next tiny delay you remove. If you run staff incentives, show dollars per hour, not minutes to bonus.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Post the total points or dollars a client can earn in the whole session, not just the next reinforcer time.
02At a glance
03Original abstract
Concurrent-chains schedules of reinforcement were arranged for humans and pigeons. Responses of humans were reinforced with tokens exchangeable for money, and key pecks of 4 birds were reinforced with food. Variable-interval 30-s and 40-s schedules operated in the terminal links of the chains. Condition 1 exposed subjects to variable-interval 90-s and variable-interval 30-s initial links, respectively. Conditions 2 and 3 arranged equal initial-link schedules of 40 s or 120 s. Experimental conditions tested the descriptive adequacy of five equations: reinforcement density, delay reduction, modified delay reduction, matching and maximization. Results based on choice proportions and switch rates during the initial links showed that pigeons behaved in accord with delay-reduction models, whereas humans maximized overall rate of reinforcement. As discussed by Logue and associates in self-control research, different types of reinforcement may affect sensitivity to delay differentially. Pigeons' responses were reinforced with food, a reinforcer that is consumable upon presentation. Humans' responses were reinforced with money, a reinforcer exchanged for consumable reinforcers after it was earned. Reinforcers that are immediately consumed may generate high sensitivity to delay and behavior described as delay reduction. Reinforces with longer times to consumption may generate low sensitivity to delay and behavior that maximizes overall payoff.
Journal of the experimental analysis of behavior, 1989 · doi:10.1901/jeab.1989.52-97