Time-dependent contrast effects in a multiple schedule of food reinforcement.

Bernheim et al. (1967) worked with pigeons in a two-part schedule.

One part gave lots of food quickly. The other gave little food slowly.

The team changed how long the rich part lasted and watched what happened when they flipped to lean.

The longer the birds stayed in the rich part, the more their pecking dropped when they hit lean.

After a short rich spell the dip was small. After a long rich spell the dip was big.

Switching from lean to rich gave a tiny burst of speed in only half the birds.

Nasr et al. (2000) later saw the same pattern in kids choosing candy. Longer nice waits made the next boring wait feel worse.

Stockhorst (1994) found people accepted leaner work if the wait to get it was short. This human test mirrors the pigeon dip: past richness sets the bar.

Eisenmajer et al. (1998) showed that even a 3-second hidden delay cuts responding. Together these papers say the same thing: value drops fast when payoff shrinks or slows.

When you alternate reinforcement rates in therapy, expect a brief slump after the better schedule ends. Keep the rich period short or add signals so the lean stretch does not feel like a shock. This tiny tweak keeps skill momentum high and cuts problem behavior that pops up when reinforcement feels thin.