Effects of concurrent response-independent reinforcement on fixed-interval schedule performance.

The team worked with pigeons on a fixed-interval 3-min schedule. Birds got grain for the first peck after 3 min passed.

At the same time, the birds also got free grain every 30 s or every 90 s. This grain did not depend on any pecking.

The researchers watched how the extra, free grain changed the normal FI scallop pattern and overall peck rate.

When free grain came every 30 s, the smooth scallop almost vanished. Birds pecked at a low, steady rate instead of waiting and then speeding up.

When free grain came every 90 s, the scallop stayed, but overall pecking dropped. The birds still paused after food, then ran faster as the interval ended.

The takeaway: the more often you give response-independent reinforcers, the more you flatten and slow the normal FI pattern.

Dove et al. (1974) saw a similar flattening under fixed-time schedules, so the finding is not a one-off. The two studies together show that any extra, non-contingent food can warp FI timing.

Davison et al. (1991) push the idea further. They showed that past schedule history and even a side water dipper can also wreck the scallop. Their work tells us that both history and current adjuncts matter, not just the concurrent food used here.

Born et al. (1974) looked at concurrent schedules too, but they changed the rate of food that still required pecking in the other key. They found temporal inhibition driven by richer alternate reinforcement. Our 1976 paper instead used completely free food, yet both studies show that what happens on the second key can pull control away from the FI. The results look different on paper, but the shared theme is clear: concurrent variables steal temporal control.

If you run FI schedules in practice, watch out for any extra reinforcers that do not require the target response. These can be tokens, edibles, or even attention delivered on a time-based plan. Too many freebies will flatten the scallop and cut responding, making the learner look like they have "lost timing" when the schedule is really just competing with itself. Check your concurrent plans and either thin the free reinforcement or fold it into the contingency so the learner can still feel the natural FI rhythm.