Within-session Changes In Responding During Concurrent Variable-interval Schedules.

F et al. watched pigeons peck two keys during concurrent VI VI sessions.

They recorded every peck and switch across the whole 60-minute period.

The team asked: does the total reinforcer rate predict how responding and switching change within the session?

Birds sped up or slowed down in step with the summed reinforcers from both keys.

Switching between keys also tracked the combined rate, not the separate rates.

One simple reinforcer-sum rule explained the whole within-session curve.

Blue et al. (1971) showed that gradual 0.5-s steps in change-over delay keep obtained rates close to programmed rates. F et al. used the same delay tactic, so their summation finding is not spoiled by abrupt delay jumps.

DeRoma et al. (2004) later zoomed into single visits and saw rapid choice shifts after each reinforcer. F et al. had already shown the broader rate-sum pattern; the 2004 paper reveals the micro-visit engine inside it.

Boutros et al. (2011) split reinforcer effects into immediate discriminative pull and long-term strengthening. Their dual view could explain why the summed rate works: each reinforcer both attracts the next response and slowly builds the overall curve F et al. tracked.

If you run concurrent schedules in a classroom or clinic, think in terms of total reinforcers per minute, not each alternative alone. A sudden dip in the combined rate will likely drop responding and increase switching, even if one schedule stays rich. Monitor the sum and you can predict within-session fatigue or resurgence before it happens.