Response acquisition with intermittent immediate and delayed conditioned reinforcement.
Keep conditioned reinforcers immediate when teaching new skills, because even short delays sharply slow learning.
01Research in Context
What this study did
Karina and team worked with rats in a lab. They wanted to see if delayed conditioned reinforcers still help animals learn new tricks.
The rats had to press a lever. Sometimes a light or sound came right after the press. Other times it came 2, 4, or 8 seconds later.
They counted how fast the rats learned the lever press when the conditioned reinforcer was immediate versus delayed.
What they found
The longer the delay, the slower the rats learned. When the light or sound came right away, rats picked up the skill quickly.
With an 8-second delay, learning almost stopped. The rats acted like no reward was coming at all.
How this fits with other research
McGonigle et al. (1982) saw the same pattern in pigeons. Shorter immediate cues worked better than longer ones. This 1982 study came first and set the stage.
Sanders et al. (1971) found monkeys and rats learned auditory tasks faster when reinforcement happened right next to the sound. Their "adjacent" timing matches Karina's "immediate" timing.
Bacon-Prue et al. (1980) showed pigeons forgot stimulus sequences after just 2-4 seconds. This memory decay helps explain why delayed conditioned reinforcers lose power so fast.
Zeiler (1968) taught rats to hold and release a bar using precise timing. His rats succeeded because the stimulus timing was tight, not delayed.
Why it matters
When you teach a new skill, deliver your praise, token, or clicker within one second. Every extra second cuts learning speed in half. If you must delay, bridge with immediate praise then give the bigger reward. This keeps conditioned reinforcers strong while you teach new behaviors.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Set a timer for one second and practice delivering your click or praise within that window every time.
02At a glance
03Original abstract
The acquisition of lever pressing by rats was studied under intermittent immediate and delayed conditioned reinforcement. Rats were exposed to two schedules operating concomitantly. One was a random-time schedule that delivered reinforcers on average of once per min. Each such reinforcer was paired with a 1-s illumination of a cue light located above the lever. The other was a tandem random-interval (RI) t₁ - s fixed-time (FT) t₂ - s schedule that controlled the rate and the delay of the illumination of the cue-light after a given lever press. For the RI, t₁ was 7.5, 15 or 30 s. For the FT, t₂ was 0, 1, 2, 4, 8, 16 or 32 s. Three rats each were exposed to each combination of RI and FT values for 40 sessions. Response rates decreased with longer response-conditioned reinforcer (S(r)) delays under the three different overall rates of S(r). At most response-S(r) delays, absolute response rates were higher with RI 15 s than with RI 30 s but both were similar to those obtained with RI 7.5 s. The results were similar to those from previous research on response acquisition with delayed primary reinforcement. These findings suggest that a previously neutral stimulus that has been correlated with primary reinforcement can reinforce new responses even when the conditioned reinforcer is intermittent and delayed from the responses that produce it.
Journal of the experimental analysis of behavior, 2013 · doi:10.1002/jeab.14