Investigating the generality of the delayed-prompt effect.
Delayed prompts alone won’t prevent errors—probe first and be ready to add error correction.
01Research in Context
What this study did
The team tested the delayed-prompt trick. They waited a few seconds before giving a prompt. The goal was error-free learning.
They ran 34 people through the drill. Each person had to pick the right card from a pair. Only four learners hit perfect scores.
What they found
Most people jumped the gun and picked wrong. Others froze and waited for help. The neat error-proof method failed most of the time.
The authors say the effect is not automatic. You must check that the learner truly knows the task.
How this fits with other research
Kisamore et al. (2016) later paired prompt delay with quick error correction. All of their kids with autism learned multiply controlled intraverbals. Adding a fix-up step seems to rescue the delay idea.
Heinicke et al. (2012) looks like a clash but is not. They showed that delaying the reinforcer by 20-40 s can block learning in kids with developmental disabilities. Castañe et al. (1993) delayed the prompt, not the payoff. Different variable, same warning: timing matters.
Syriopoulou-Delli et al. (2012) also saw mixed results when they swapped delayed prompt for a delayed error-correction tactic. Again, the tool worked for some kids and flopped for others. The pattern fits: delay helps only when it matches the learner’s skill.
Why it matters
Do not trust delayed prompts to erase errors on their own. Run a quick probe first. If the learner starts guessing, add immediate error correction or shorten the delay. Check again next session—what worked today may fail tomorrow.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Before starting a prompt-delay program, run five probe trials with no prompt; if errors top 40%, add immediate error correction.
02At a glance
03Original abstract
Two experiments were conducted in an attempt to replicate the findings of Touchette's classic 1971 experiment. The results obtained were more variable than those obtained by Touchette. Four of 34 subjects learned the Horizontal E discrimination without errors. Ten subjects preempted the red prompt stimulus erroneously. Twelve subjects failed to preempt. Among those who failed to preempt, at least six had not learned the target discrimination. There were also five subjects who failed to preempt who, on subsequent tests, demonstrated that they had learned the target discrimination. For these subjects, the onset of the prompt appeared to grant "permission to respond." An attempt is made to account for the differences in the current results and those of Touchette (1971). We also compare our results to the usually successful application of the delayed prompt to education and training problems. We note that the procedures used in these successful applications are often quite different from those in the original Touchette experiment and our own.
Research in developmental disabilities, 1993 · doi:10.1016/0891-4222(93)90036-j