Effect of varying the duration of grain presentation on automaintenance.

The team worked with pigeons in an automaintenance setup. A key light came on for a few seconds. Then grain arrived no matter what the bird did. The researchers changed how long the grain hopper stayed open. They tried short, medium, and long durations. Sometimes the durations stayed the same for many sessions. Sometimes they switched within the same session.

The goal was to see if longer grain time would make the birds peck faster or more often.

The results were messy. Longer grain did not steadily boost pecking. The only clear hint was a tiny drop in peck latency when short and long durations mixed in the same session. Even that effect was small and did not repeat every time.

Overall, grain length barely mattered for keeping the key-peck alive.

Frost et al. (1996) looked at pellet size instead of duration. Bigger pellets gave faster, stronger pecking every time. Their positive result shows that reinforcer magnitude can matter, but how you change it counts. Size changes worked; duration changes did not.

Schmidt et al. (1969) also tested duration years earlier. They used a discrimination task, not automaintenance. When they cut duration from 6 s to 2 s and added a hopper light, pigeons showed a clear peak shift. Duration mattered there because the birds had to tell the durations apart to earn food.

Staddon (1970) saw duration effects too, but on a fixed-interval schedule. Long grain stretched the post-reinforcement pause. The schedule context made duration important. In pure automaintenance, that context is missing, so duration fades out.

If you run autoshaping or pairing procedures with clients, do not assume longer reinforcer time will strengthen behavior. Check first whether the task gives the learner a reason to notice the reinforcer details. When the only job is “sit and watch,” duration probably will not drive learning. Save your effort for variables like stimulus salience or immediacy instead.