Behavioral contrast with fixed interval and low-rate reinforcement.

The team tested pigeons on two low-rate schedules: fixed-interval and DRL. Birds pecked a key that sometimes paid off after a set time or only if they waited long enough between pecks.

The twist: a colored light told the bird whether the next round would pay or not. They switched the meaning of the colors to see if response rates still climbed when the ‘no-pay’ color came back.

Contrast still happened. Key pecks rose when the extinction cue returned, even though the schedule itself never asked for fast responding.

This showed that contrast is not just about how quickly you must respond to get food. The sight of a formerly rich cue going dry can push rates up on any schedule.

Bloomfield (1967) watched the same jump in pecking and added that the spike can linger after pay returns. The 1961 study set the stage; the 1967 follow-up mapped how long the hangover lasts.

Dixon (2014) later offered a new rule book. His math model says contrast comes from hidden, schedule-induced behaviors competing for time. This 2014 theory now covers the 1961 puzzle: even slow schedules can spark fast side behaviors that fill the gap.

Parsons et al. (1981) seemed to disagree at first. They saw flat data until they removed accidental pay for early pecks. Once the setup was clean, contrast popped out, backing up the 1961 claim that the effect is real but fragile.

If you run multiple-schedule sessions or use colored cues for different tasks, know that taking reinforcers away from one cue can inflate behavior under another, even when the remaining schedule favors slow, careful responding. Check your data after you thin or remove reinforcement from any component; a jump elsewhere may be contrast, not skill gain. To keep rates steady, fade reinforcement gradually and watch for accidental payoff that could mask or mimic the effect.