A yoked-chamber comparison of concurrent and multiple schedules: the relationship between component duration and responding.
Matching on multiple schedules can fool you—short turns, not true choice, create the curve.
01Research in Context
What this study did
Researchers compared two ways to arrange reinforcement schedules.
One was concurrent schedules where two choices are always present.
The other was multiple schedules where choices appear for short turns.
They used pigeons in matching chambers to see how each setup shaped response patterns.
The key twist was keeping component durations very short in the multiple schedule.
What they found
Birds showed matching in both setups, but for different reasons.
On concurrent schedules, matching came from true choice between options.
On multiple schedules, matching looked real but was driven by the brief turns.
Short components forced quick switches, creating a fake match to reinforcement rates.
How this fits with other research
Lobb et al. (1977) later confirmed the finding. They showed even stronger undermatching on multiple schedules and proved the result stays the same across contexts.
Jarrold et al. (1994) moved the idea into a classroom. Students doing math under concurrent schedules also undermatched unless teachers added extra supports like changeover delays.
Marcucella et al. (1978) widened the warning. They showed that even on concurrent schedules, simply signalling when reinforcement is ready can wreck matching, just like short turns do.
Why it matters
When you run multiple-schedule probes, do not trust perfect matching as proof of preference. Short components can fake the curve.
Before you call a child’s data a true choice, check how long each option stays available.
If you need honest preference data, use concurrent schedules with long, stable components or add safeguards like changeover delays.
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.
Lengthen each component to at least two minutes before you trust matching data.
02At a glance
03Original abstract
Two experimental chambers were electrically connected so that the component selected by a pigeon confronting concurrent variable-interval schedules in one chamber could be successively presented as a multiple schedule to a second pigeon in the other chamber. Component duration was regulated by the use of a changeover delay, the value of which was systematically varied between 0 and 30 sec. It was found that the relative local response rates on the preferred key (absolute response rate to that component divided by the sum of the absolute response rates during both components) tended to increase with increasing component durations for the birds in the concurrent chamber, but decreased for the birds in the multiple chamber. These data support the interpretation that there are fundamental differences in the mode of responding to multiple and concurrent schedules. Based on these findings, it was concluded that previous demonstrations of matching on multiple schedules do not establish that response allocation is controlled by a process equivalent to that found on choice paradigms. It now appears that matching on multiple (but not concurrent) schedules is a consequence of selecting short component durations. The implications of these data for Herrnstein's (1970) and Rachlin's (1973) formulations of the relationship between multiple and concurrent schedules are examined.
Journal of the experimental analysis of behavior, 1974 · doi:10.1901/jeab.1974.22-21