An analysis of coordinated responding of pigeons
Requiring two learners to act together boosts joint responding, but too many choices breaks the sync.
01Research in Context
What this study did
Katz and team worked with pairs of pigeons. Each bird had its own key.
The birds only got food if both pecked within one second of each other.
They tested this with two, four, or eight keys per bird to see if choice load mattered.
What they found
When food depended on both birds pecking together, the pairs synced up fast.
Adding more keys weakened the sync. With eight keys each, the birds often missed the one-second window.
Fewer keys kept the teamwork tight.
How this fits with other research
Garcia et al. (1973) showed pigeons match their response rate to the payoff rate. Katz adds a twist: the payoff now hinges on two birds acting as one.
Calamari et al. (1987) studied interdependent links in a chain schedule. They saw early bursts of pecking. Katz sees the same burst, but only when the key set is small.
Fahmie et al. (2013) found that past delays can override current payoff rates. Katz echoes this: extra keys act like added delays, diluting the power of the coordination rule.
Why it matters
If you run social skills groups, limit the choices. Two kids sharing one iPad turn-taking game will sync better than eight kids with eight tablets. Start small, then add options only after the pair shows smooth teamwork.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick two learners, give them one shared toy, and reinforce only when both use it within three seconds.
02At a glance
03Original abstract
Experimental analyses of coordinated responding (i.e., cooperation) have been derived from a procedure described by Skinner (1962) in which reinforcers were delivered to a pair of subjects (a dyad) if both responded within a short interval, thus satisfying a coordination contingency. Although it has been suggested that this contingency enhances rates of temporally coordinated responding, limitations of past experiments have raised questions concerning this conclusion. The present experiments addressed some of these limitations by holding the schedule of reinforcement (Experiment 1: fixed ratio 1; Experiment 2; variable interval 20 s) constant across phases and between dyad members and by varying, in different conditions, the number of response keys (one to three) across which coordination could occur. Greater percentages of coordinated responding occurred under the coordinated-reinforcement phases than under independent-reinforcement phases in most conditions. The one exception during the one-key condition of Experiment 1 appeared to be a consequence of variability introduced by the independent-reinforcement phase procedure. Furthermore, coordination percentages decreased with increasing response options under both schedules. These results confirm and extend the finding that coordination contingencies control higher rates of temporally coordinated responding than independent-reinforcement contingencies do.
Journal of the Experimental Analysis of Behavior, 2024 · doi:10.1002/jeab.899