Response latency as a function of reinforcement schedule.

Researchers looked at how often reinforcement is given changes how fast pigeons let go of a key. They used a single-case lab setup. The birds had to release a key quickly after a light came on.

Different schedules of food reward were tested. The team watched how long each bird waited before letting go.

More frequent food produced faster key releases. Latencies were also steadier, with less scatter between trials.

When rewards were thin, the birds hesitated longer and the times bounced around more.

Glover et al. (1976) built on this idea. They showed that after food, pause length and running rate are controlled by separate parts of the schedule. C et al. found latency changes; J et al. explained why pause and speed can move apart.

Blough (1992) later confirmed that probability of reward, not how big or long it lasts, drives reaction time. This sharpens the 1962 takeaway: frequency matters, size does not.

Gaucher et al. (2020) extended the same rule to children with autism. Under a DRL schedule, kids who could time their responses had higher IQ and language scores. The latency principle jumps from lab pigeons to clinic learners.

When you thin a schedule, expect brief spikes in latency or pause. Use dense reinforcement at first to lock in quick, consistent responses. Then fade slowly while watching both latency and post-reinforcement pause. If you run DRL for rate reduction, pre-check the learner's timing and language skills; quick adjustment is harder for kids still learning to wait.