The effect of signaled reinforcement on rats' fixed-interval responding.
The learner’s exact response speed right before your praise or click decides whether that signal will speed them up or slow them down next time.
01Research in Context
What this study did
Reed (2003) worked with rats on a fixed-interval schedule. Every few minutes, the first lever press after time ran out earned food.
Half the time a 500-ms tone or light came on with the food. The team watched how this tiny signal changed the rats’ next burst of pressing.
What they found
The same beep did opposite things. If the rat had just pressed fast, the signal made the next burst even faster.
If the rat had pressed slowly, the same beep slowed the next burst. The moment-to-moment rate right before food decided the direction.
How this fits with other research
LeFrancois et al. (1993) showed that only the last schedule, not old history, sets FI rate. Phil adds that even a 500-ms signal follows the same local rule.
Bland et al. (2016) later saw the same bidirectional pattern in DRA: a signal can make alternative reward more powerful and make problem behavior stick around longer.
Laugeson et al. (2014) looks like a clash — they gave pigeons free food with a signal and saw no change in extinction. The difference is timing: Phil’s signal hit right after the rat’s own response, while A’s signal came before free food that had nothing to do with pecking.
Why it matters
Your “good job” or click is not neutral. If the learner just worked hard, the praise will likely push the next burst higher. If they were barely responding, the same praise can suppress the next try. Watch the immediate pace before you deliver any conditioned reinforcer and adjust your timing or wording to get the lift you want.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Count the learner’s responses in the final five seconds before you give praise; if the rate is high, deliver the praise right away—if it is low, wait one beat or skip the signal that round.
02At a glance
03Original abstract
Four experiments examined the effect on rats' response rate of presenting a brief (500 ms) stimulus simultaneously with the delivery of food on fixed-interval (FI) schedules. In Experiment 1, reinforcement signals that were spatially diffuse (both tones and lights) elevated rates of responding, but responding was attenuated by localized visual stimuli. The remaining experiments examined the signal-induced potentiation of responding. In Experiment 2, a tone reinforcement signal potentiated response rates on an FI schedule, but attenuated response rates on a variable-interval (VI) schedule. This difference was obtained even though the overall rate of responding was equated on the two schedules before the introduction of the signal. Signal-induced potentiation of responding occurred over a range of FI values employed in Experiment 3. In Experiment 4, presenting a reinforcement signal when high local rates of response had occurred immediately before reinforcement resulted in potentiated rates of responding on an FI schedule. The opposite effect on response rate occurred when the reinforcement signal followed only low local rates of response. These results indicate that a variety of factors influence the effects of a reinforcement signal. They imply, however, that the local rate of response at the time of reinforcement is a key factor in establishing the nature of the signaling effect.
Journal of the experimental analysis of behavior, 2003 · doi:10.1901/jeab.2003.79-367