MULTIPLE BASELINE INVESTIGATION OF STIMULUS FUNCTIONS IN AN FR CHAINED SCHEDULE.

GOLLUMIGLER (1964) compared two schedules that looked identical to the bird. Both had three fixed-ratio links. The only difference was the lights. In the chained schedule, each link had its own color. In the tandem schedule, the same color stayed on the whole time.

Pigeons pecked for food. The team recorded how fast they pecked in each link. They wanted to see if the color changes alone could change behavior.

Birds slowed down in the early links when colors changed. They paused more and pecked less. In the tandem schedule with one steady color, they kept a steady pace.

The color cues—not the food rules—controlled the response pattern.

Henton (1972) seemed to disagree. When pigeons could choose between chained and tandem terminal links, they picked each side equally. Same birds, same lights, no preference. The difference: W measured choice, not speed. Choice tests look at delayed primary food, while rate tests catch moment-by-moment stimulus control.

Horner (1971) supports the 1964 story. Repeating an early color in later links did not speed birds up. Only moving closer to food raised rate. Proximity, not the light itself, drives the pattern.

Wacker et al. (1985) and Reed et al. (1988) confirm the downside of chaining. Given a concurrent choice, birds reliably avoid chained schedules. The 1964 rate dip turned into a long-term preference for simpler, tandem-like schedules.

When you build multi-step programs, each new stimulus can act like a chained link. Early tasks may see pausing or avoidance even if the final reinforcer stays the same. You can reduce the effect by keeping stimuli steady until the learner reaches the terminal step, or by adding extra prompts near the start. Check response rate across links—if it drops early, the stimulus change itself could be the culprit, not the work requirement.