Latency and frequency of responding under discrete-trial fixed-interval schedules of reinforcement.

Dove et al. (1974) watched how fast and how often pigeons pecked under fixed-interval schedules. They used discrete-trial FI. Each trial ended after one response or after time ran out.

The team tracked two things: latency (time until the bird pecked) and frequency (how many pecks inside the interval).

Latency always dropped as the interval moved on, making a smooth scallop no matter the FI length. Response frequency did not follow a single rule; it changed when the schedule details changed.

In short, latency gave a steady time picture, while frequency bounced around.

Cicerone (1976) later pitted FI against variable-interval and showed FI gives the clearest time control, backing the steady latency scallop D et al. saw.

McSweeney et al. (1993) explained why the scallop happens: the post-reinforcement pause tracks the last interfood time, a process D et al. captured as latency.

Greene et al. (1978) found humans only show the scallop when a digital clock is in the room; without it, the pattern falls apart. This extends D et al. by showing the latency rule needs the right cues outside the animal lab.

Mattson et al. (2024) reviewed 27 studies and say latency is now a practical choice in functional analysis when frequency is hard to count, putting the 1974 lab trick into everyday ABA toolkits.

When you run or interpret FI-based programs, measure latency if you want a number that stays stable across settings. Use it in functional analyses for low-rate or unsafe behaviors where counting every response is tough. Pair the measure with clear timing cues (like a visible countdown) so clients show the expected scallop instead of break-and-run. This old pigeon finding still sharpens your data decisions today.