Concurrent performances: rate and accuracy of free-operant oddity responding.
Reinforcement on an alternate key directly dims—or boosts—oddity accuracy, so watch concurrent rates as closely as you watch the target task.
01Research in Context
What this study did
The team set up a free-operant oddity task. Adults without disabilities pressed keys to pick the one shape that did not match.
Two keys sat side-by-side. One key paid off on a variable-interval (VI) schedule. The other key gave the oddity task. The VI rate changed across phases while everything else stayed the same.
What they found
Oddity accuracy went up when the VI key paid off less. Accuracy stayed high while the VI key paid off more.
Inside a multiple schedule, VI components always beat extinction components for accuracy. The result held even when the adults switched keys at the same speed.
How this fits with other research
Schroeder et al. (1969) used the same lab and the same concurrent VI set-up, but shaped eye movements instead of oddity. Both studies show that changeover delays keep behavior in place and let reinforcement rate steer the response.
Amore et al. (2011) later moved the same multiple-schedule logic to children with intellectual disability. Rich-lean contrast made their behavior more resistant to disruption, echoing how VI-rich segments lifted oddity accuracy here.
Lozy et al. (2019) looks like a clash at first: they found that low-preferred reinforcers still drove high compliance in single-operant tasks. The key difference is task type. Their compliance task ran alone, while the 1972 oddity task ran next to a paying side key. Concurrent competition, not preference, ruled accuracy.
Why it matters
You now have lab proof that side reinforcement can drag down discrimination accuracy, even when the learner stays busy. If a client splits time between work and a reinforcing ‘break’ option, cut the break rate or enrich the work rate to bring accuracy back up. Check accuracy across rich and lean components within the same session; the contrast will tell you if the reinforcement balance is right.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Track accuracy in rich and lean components within the same session; lower the side-task payoff if accuracy dips.
02At a glance
03Original abstract
In pigeon's oddity performances, maintained by variable-interval reinforcement of pecks on the odd key of three keys in a triangular array, accuracy and response rate varied inversely with the rate of variable-interval reinforcement scheduled concurrently for pecks on a fourth, spatially isolated key. But when variable-interval and extinction components alternated in a multiple schedule for pecks on the spatially isolated key, oddity accuracy was greater during variable-interval components than during extinction components. Oddity response rate was not affected systematically by the alternating components. Changeovers between the oddity keys and the spatially isolated key were frequent during variable-interval components; responding occurred almost exclusively on the oddity keys during extinction components. This difference in performance during the two components was eliminated by arranging stimulus-correlated variable-interval reinforcement in the multiple schedule on the spatially isolated key: a stimulus was presented in the variable-interval components only when reinforcement became available, thereby reducing responding on this key to near-zero levels in both components while maintaining the variable-interval reinforcement. The effect of the multiple-schedule components on oddity accuracy was not altered, however, and thus apparently depended directly on concurrent reinforcement and not on differential sequential properties of concurrent responding during the two components.
Journal of the experimental analysis of behavior, 1972 · doi:10.1901/jeab.1972.17-25