Briefly delayed reinforcement: An interresponse time analysis.

The team worked with pigeons on a variable-interval food schedule. They added a tiny, half-second delay between each peck and the grain hopper.

Sometimes the delay came with a brief light cue. Other times it arrived with no signal at all. They counted how fast the birds pecked and how long they waited between pecks.

No-signal delays made the birds peck faster and bunch their pecks closer together. A tiny signal wiped out that speed-up and even slowed the birds down.

The pattern flipped when the delay was moved after the food instead of after the peck. Results were messy, showing that timing matters at the half-second level.

Sievert et al. (1988) used the same half-second no-signal delay and showed the speed-up happens because the bird now treats a quick burst, not a single peck, as the response that earns food. Wolchik et al. (1982) spotted the burst; L et al. explained why it forms.

Guest et al. (2013) repeated the effect but only under interval schedules. When they switched to ratio schedules the burst vanished. The lesson: brief delays hurt interval performance more than ratio performance.

Hamm et al. (1978) looked at many delay lengths and saw rates rise with short delays then fall with long ones. A et al. zoomed in on the rise side, confirming the early bump in the curve.

If you run DRL, VI, or any schedule where you pause to deliver tokens, check your delay. Even the time it takes to open a container or tap a tablet can accidentally reward rapid, sloppy responses. Add a quick signal—like a brief beep or color change—to keep rates calm and accurate.