Second-order schedules: a comparison of chained, brief-stimulus, and tandem procedures.
Brief stimuli that are paired with food act like mini-reinforcers and speed up early responding in big schedules.
01Research in Context
What this study did
Pomerleau et al. (1973) tested seven ways to link brief lights or sounds to food delivery. They used pigeons pecking a key under second-order schedules. Each schedule had two parts. The birds first finished a small fixed-interval unit. A brief stimulus then flashed. After several units, real food arrived.
The team compared chained, tandem, and brief-stimulus versions. In chained, the brief cue also told the bird what came next. In tandem, the cue had no extra information. In brief-stimulus, the cue only paired with food. All other details stayed the same.
What they found
Food-paired brief stimuli made the pigeons peck faster early in each unit. The response curve looked like a small fixed-interval scallop. Chained and tandem cues without direct food pairings did not speed up responding. The brief-stimulus version created the strongest early acceleration.
This pattern showed the food-paired cue worked as a strong conditioned reinforcer. It pushed responding even when it gave no information about what came next.
How this fits with other research
Gibbon (1967) ran a simpler second-order FI with brief conditioned reinforcers. Pomerleau et al. (1973) built on that base by adding schedule comparisons. The earlier paper proved brief cues could maintain FI patterns. The 1973 paper showed which cue type does it best.
Downing et al. (1976) later varied the order of food and stimulus deliveries. They found random mixing made the cue even stronger. Together, the two studies tell us both cue-food pairing and unpredictable sequencing boost conditioned reinforcement.
Gardner et al. (1977) swapped food for cocaine while keeping the brief stimulus. Cocaine raised rates more, but removing the cue flattened the FI pattern. This extends F et al. by showing the brief-stimulus effect holds across different reinforcer types.
Why it matters
You can turn any brief event into extra power for your reinforcement system. Pair a click, flash, or token with the primary reinforcer. Deliver it right after the target response. You should see earlier and faster responding within large schedules. This matters when you stretch ratio or interval requirements with clients. One click now can save many food bites later.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pair your token drop with immediate praise and a click; watch if the first responses in each ratio come faster.
02At a glance
03Original abstract
Pigeons were exposed to seven types of two-component schedules, each component a 2-min fixed-interval schedule. Food presentation occurred at the completion of the second component under all conditions. The seven types of schedules were: (1) a chained schedule in which completion of the first component produced the discriminative stimulus associated with the second component; (2) a chained schedule to which was added the brief presentation of a food-paired stimulus at the completion of the first component; (3) a chained schedule to which was added the brief presentation of a stimulus not paired with food at the completion of the first component; (4) a multiple schedule in which food presentation occurred at the completion of both components; (5) a tandem schedule in which completion of the first component initiated the second component, with no changes in exteroceptive stimuli; (6) a food-paired brief-stimulus schedule in which the brief presentation of a food-paired stimulus was made at the completion of the first component and no other changes in stimuli occurred; and (7) a brief-stimulus schedule in which the brief presentation of a stimulus not paired with food was made at the completion of the first component and no other changes in stimuli occurred. Positively accelerated patterns of responding developed in the first component under three conditions: (1) the chained schedule with the added food-paired brief stimulus; (2) the multiple schedule; and (3) the food-paired brief-stimulus schedule. Response rates were low in the first component, with few instances of positively accelerated patterns, under two conditions: (1) the chained schedule; and (2) the chained schedule with the added nonpaired brief stimulus. The results suggest that a briefly presented food-paired stimulus may function as a more effective conditioned reinforcer than does the presentation of a discriminative stimulus.
Journal of the experimental analysis of behavior, 1973 · doi:10.1901/jeab.1973.20-447