DRAGENwiki · Article
S-Delta
Customize your experience
Sources checked. Clinical review pending.
An S-delta (S∆) is something a person can see, hear or notice that signals a certain action will not pay off right now, because that action has not been reinforced when the signal was there. [1]
S-deltas in 60 seconds
- It means "not now" for one action. An S-delta goes with reinforcement being unavailable for a response in its presence. [1]
- It is the partner of an SD. A discriminative stimulus (SD, also called S+) signals that a response can pay off. An S-delta (also called S−) signals that it will not. [1]
- Its role can change from moment to moment. In a set of pictures, the same picture can be the right answer on one trial and a wrong one on the next. [2]
- Learning can happen with few errors. Errors were not needed to learn a discrimination when the wrong choice started out very different from the right one. [3][4]
- Not-now signals help in daily life. In functional communication training (FCT), clear signals for yes and not now kept asking strong while giving caregivers breaks. [5][6]
Jump to: What an S-delta is · Every SD has a partner · Learning with few errors · Not-now signals · When a signal is not enough · DRAGEN's rules · Example · Mistakes · Words to know · Need a PDF? · More from DRAGEN · Sources
This page is free, like the rest of DRAGENwiki. It is the partner page of discriminative stimulus, and it builds on stimulus control and functional communication training.
What an S-delta is
The short version. An S-delta is a signal that one certain action will not pay off right now. A closed sign on a shop door signals that pulling the handle will not get you inside. A busy tone signals that talking into the phone will not reach anyone. You learn these signals because, when they were there, that action did not work.
The research definition. One review describes it this way: "an S-delta (S∆, also referred to as S−) is associated with the unavailability of reinforcement for a response in its presence". [1] Reinforcement here means a payoff that makes the action more likely to happen again.
It belongs to one action, not to the person. The definition is about "a response", one action. [1] A closed sign says nothing about you. It says only that pulling the door will not work right now. Other actions still can: you can knock, call, or come back later. In the same way, a signal that asking for a snack will not be answered is not a signal that asking for help will not be answered.
The S-delta can simply be the SD gone. In one early FCT study, as a later paper describes it, a picture card hung on the wall signaled that asking would be answered. When the card was taken down and put out of sight, asking was not answered. The missing card worked as the S-delta. [6] In that later paper's own study, a therapist wore a yellow wristband to signal that asking would be answered, and its absence signaled that it would not. [6]
Every SD has a partner
Right and wrong, at that moment. In a listening task, an adult says a word and the learner chooses a picture. A teaching guide explains that the spoken word "establishes one of the comparison stimuli as the correct one (i.e., discriminative stimulus) and the others as distracters (i.e., S-deltas) at that particular moment." [2]
The same picture can play both roles. The guide adds: "Each of the comparison stimuli functions as the discriminative stimulus or distracter on a given trial." [2] So a picture of a cup is an S-delta when the adult says dog, and the SD when the adult says cup. The right choice depends on what was asked. That is a conditional discrimination.
How different the partners look matters. Researchers separate two ideas. Disparity is how different the SD and the S-delta are from each other. Salience is how much the SD stands out from its background. [1]
Make the difference easy first, then fade. The same review suggests arranging conditions that "highlight the critical features of the S+ and S− so they are more salient to the observer". Then the extra help is faded step by step, so the targeted SD takes over. [1]
Learning with few errors
Errors are not required. In a classic lab study with pigeons, responses to the S− (the errors) were not needed to learn a color discrimination. [3] At first, the wrong key was dimmer and shown for less time than the right one. Step by step, it was made brighter and longer, until the only difference left was its color. Training that started this way, early on, produced no errors. [3]
Errors had side effects. Errors did happen when training started late, or when the two keys differed only in color the whole time. Only the birds that learned with errors showed what the paper called emotional responses to the S−, faster responding to the S+, and occasional bursts of responses to the S−. [3]
FCT research draws on this. One FCT paper notes that basic research has shown that "gradual introduction of the negative stimulus (SΔ) often results in accurate discrimination without response bursting or the corollary emotional responses that often occur when EXT is abruptly introduced". [6] EXT stands for extinction: the response no longer gets its payoff. [5]
With children, too. In an older study with children, as a later paper describes it, one group learned to choose a circle over ellipses with size fading. [4] "The height of the incorrect stimuli (ellipses) was much lower than the circle at first and became more similar to the height of the circle over trials." [4] Compared with children who started with small differences, the children taught with size fading "made fewer errors, ultimately maintained higher accuracy", and could tell apart smaller differences in size. [4]
In plain words. Start with a wrong choice that is easy to tell apart, then make it more like the right one, one small step at a time. The pages on errorless learning and prompting go further.
Not-now signals in daily life
The problem FCT runs into. FCT teaches a person another way to get the same kind of payoff that a problem behavior was getting. [5] At first, every request is answered. One guide writes: "we strongly recommend that the communicative response be reinforced on a CRF schedule initially." [5] CRF means continuous reinforcement: every request gets its payoff. But caregivers have other things to do, such as cooking or caring for other children. [5] If answers suddenly become rare, "the response is likely to undergo extinction, setting the stage for the reemergence of problem behavior". [5] So FCT plans usually include schedule thinning: making answers less frequent, step by step, so requests do not come faster than caregivers can answer. [5][7]
Red card, white card. In one study, as the guide describes it, people learned to tell when asking would not be answered. Each period had its own card: "a red card viewable during reinforcement periods, and a white card viewable during extinction periods". [5] The guide reports: "This stimulus control procedure was highly effective when reinforcement for the communicative response was only available for 20% of the time." [5] It kept a strong link between asking and getting, and it still gave caregivers a break during the not-now periods. [5]
Short not-now periods first. A later paper describes how that study began: 45 seconds when asking was answered, then 15 seconds when it was not. Through seven steps in between, it reached 1 minute of yes and 4 minutes of not now. [7] The final schedule "allowed caregivers intervals of 4 to 8 min at a time during which no reinforcer deliveries were programmed, thus allowing them time to complete other routines." [7]
A wristband and a rule. In another study, a therapist wore a yellow wristband when asking would be answered and took it off when it would not. At the start of each session, the therapist said the rule out loud. [6] When these signaled schedules began, "the functional communication response came under rapid discriminative control, and problem behavior remained at near-zero rates." [6] The authors suggest that signals like these might also help results carry over from one setting or therapist to the next. [6]
Everyday signals. In one study, as a later paper describes it, two children were taught to ask for attention when an adult was not busy, such as sitting and doing nothing, but not when the adult was busy, such as cooking. [7] One child had a favorite item during busy times. The other was taught to check by asking, Are you busy? [7] When the final package was used with two pairs of busy and not-busy activities, "treatment effects rapidly generalized to several other novel pairs." [7]
Arranged or natural signals? A study with three participants compared signals a therapist arranged, like colored cards, with signals that happen naturally, like what the therapist is doing. Two of the three learned when to ask more quickly with the arranged signals. Two of the three also carried the skill to new settings better with arranged signals. [7]
When a signal is not enough
Expect the old behavior to come back for a while. The guide is direct: "Problem behavior should be expected to occur during the reinforcement thinning process, regardless of the strategy used." [5]
Resurgence is common. One review looked at 25 uses of schedule thinning in FCT and found resurgence, the old behavior coming back, in "19 of the 25 (76%) applications of reinforcement schedule thinning". [8] All 25 used signaled periods for yes and not now. [8] In some cases, the behavior came back at higher levels than before treatment. [8] When the behavior had been getting a break from demands, it came back at least once in every case, 8 of 8. [8]
One method took away the card. In 7 of the 25 uses, researchers removed the request card during not-now periods. [8] DRAGEN does not take away a person's way to communicate. See the rules below.
Shorter first waits might help. In the arranged-signal study, two of the three participants showed resurgence. [7] The authors wrote: "Had we started with much shorter EXT components, it is possible that resurgence would have been mitigated (or even prevented) for these two participants." [7]
Give the wait something to do. The guide's advice pairs clear signals with "making alternative reinforcers available during delay or extinction periods". [5] In plain words: during not now, offer something else worth doing.
Make not now honest and short
DRAGEN's rules for not-now signals. These are page rules from the four pillars, not findings:
- Not now must be honest. When the not-now signal ends, yes comes back, as promised.
- Needs are never not now. Help, the bathroom, water, pain, feeling sick, safety and stop get an answer right away, signal or no signal.
- A no is not put on hold. Not now covers a want. A person's no, refusal or protest gets a response.
- Never take away a way to communicate. A signal may cover one request. It never removes a person's voice, device or cards for needs and for no.
- Start short. Lengthen not-now periods slowly, on a written plan, with data.
- Fill the wait. Offer something worth doing during not-now periods.
- Teach checking. Teach the learner a way to check the signal, such as asking Are you busy?
- A BCBA plans and watches the data. A BCBA plans schedule thinning and reviews the data as it goes.
A no is information. Saying no, asking a question, or taking back a yes tells you about the task, the timing or the person's needs. It is not noncompliance, and it is not a treatment target.
Safety is a separate question. If someone could get hurt right now, handle safety first. Do not use safety as a reason to stop listening to dissent.
This page is general information, not medical or legal advice. A BCBA reviews every clinical decision.
A made-up example: Learner A
Learner A is a made-up child who used to scream to get a parent's attention while the parent cooked dinner.
What the team saw. An assessment planned by a BCBA suggested the screaming got attention. With FCT, Learner A learned to tap the parent's arm and say play with me. The parent answered every time. Soon Learner A was asking every minute, and dinner was not getting made.
What they set up. The parent wore a green wristband when play with me would be answered, and put it in a pocket when it would not. At the start, the parent said a simple rule: When the band is on, ask me and I will play. When it is off, I am cooking, and I will play again soon. Not-now periods started at about 15 seconds. Learner A had a favorite puzzle and a picture schedule that showed when play would come back.
What did not change. Asking for help and asking to use the bathroom were answered right away, band or no band. When Learner A pushed the puzzle away and said no puzzle, the parent offered something else and wrote it down as information about what Learner A wanted, not as a problem.
What happened in week two. As the not-now periods got longer, screaming came back for a few days. The BCBA had told the family this might happen. They stepped back to shorter waits, then lengthened them more slowly.
What happened next. In this made-up story, Learner A learned to look for the band and to ask Are you busy? before tapping. Real learners differ, and a BCBA reviews every plan.
Common mistakes
- "An S-delta means no forever." It signals that one response will not pay off while the signal is there. [1]
- "A wrong picture is always wrong." The same picture can be the right answer on one trial and a distracter on the next. [2]
- "People must make errors to learn." Discriminations have been learned with few errors when the wrong choice started out very different. [3][4]
- "Stretch the wait fast." Plans usually lengthen not-now periods step by step, and one study's authors thought shorter first waits might have prevented resurgence. [5][7]
- "A clear signal is enough." The old behavior often comes back during thinning, even with signals. [5][8]
- "Everyday signals always work better." In one study, arranged signals led to faster learning for two of three participants. [7]
Words to know
- S-delta (S∆): a stimulus that goes with reinforcement being unavailable for a response in its presence. Also called S−. [1]
- Discriminative stimulus (SD): a stimulus that goes with reinforcement being available for a response in its presence, and not in its absence. Also called S+. [1]
- Distracter: in a set of pictures, a choice that is the incorrect response on a given trial. [2]
- Errorless discrimination training: teaching a discrimination with few or no responses to the S-delta, often by bringing the S-delta in gradually. [3][6]
- Multiple schedule: signaled periods when a response is reinforced (SD) and when it is not (S-delta). [6][7]
- Extinction (EXT): reinforcement no longer follows a response. In a multiple schedule, the extinction component is the not-now part. [5][7]
- Schedule thinning: making reinforcement less frequent step by step, for example by lengthening not-now periods. [5][7]
- Functional communication training (FCT): teaching another way to get the same kind of payoff a problem behavior was getting. [5]
- Resurgence: an old behavior coming back when reinforcement for the new response is thinned. [8]
- Observing response: an action that lets a person check the signal, like asking Are you busy? [7]
- Disparity and salience: how different the SD and the S-delta are from each other, and how much the SD stands out from its background. [1]
Need a PDF?
Create it here, for free: the DRAGEN app. Use it at your next session or family meeting.
Or print one we already made, also free:
- Ask cards (PDF)
- Replacement plan (FCT) (PDF)
- Picture schedule (PDF)
- First-then board (PDF)
- Treatment integrity checklist (PDF)
- Every free PDF
More from DRAGEN
- Lesson: How to teach a child to ask instead: functional communication training (FCT)
- Lesson: How to use a first-then board and a picture schedule
- Lesson: How to help without doing it for them: prompting and prompt fading
- Lesson: How to tell the four functions of behavior apart
- Free app for parents: DRAGENparent
- BCBA care: DRAGENtherapy
- Books: DRAGENbooks
- Talk with Chase: Contact
- Home: dragen.cloud
Related DRAGENwiki pages: discriminative stimulus · stimulus control · stimulus discrimination · conditional discrimination · functional communication training · mand · motivating operations · four-term contingency · errorless learning · prompting · faulty stimulus control · functional analysis · right to say no · procedural fidelity
What remains unknown
- Small studies. The wristband study and the arranged-signal study each had three participants. The resurgence review covered 25 uses of schedule thinning. [6][7][8]
- Which signal suits whom. Arranged signals helped two of three people learn faster in one study. It is not yet clear which signals work best for which learners. [7]
- How the wait feels. These studies measured asking and problem behavior, not how the person felt during not-now periods. [6][7][8]
- Second-hand evidence. The red and white card study, the busy and not-busy study, the size-fading study and the 25 uses are known here only as other papers describe them. [4][5][7][8] The pigeon study is cited from its abstract. [3]
- This page was last checked on 2026-11-12.
Sources
- Halbur, M. E., Caldwell, R. K., & Kodak, T. (2021). Stimulus control research and practice: Considerations of stimulus disparity and salience for discrimination training. Behavior Analysis in Practice, 14(1), 272–282. https://doi.org/10.1007/s40617-020-00509-9 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC7900359/)
- Grow, L., & LeBlanc, L. (2013). Teaching receptive language skills: Recommendations for instructors. Behavior Analysis in Practice, 6(1), 56–75. https://doi.org/10.1007/BF03391791 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC3680153/)
- Terrace, H. S. (1963). Discrimination learning with and without "errors". Journal of the Experimental Analysis of Behavior, 6(1), 1–27. https://doi.org/10.1901/jeab.1963.6-1 (abstract read at https://pmc.ncbi.nlm.nih.gov/articles/PMC1404228/)
- Saunders, K. J., Hine, K., Hayashi, Y., & Williams, D. C. (2016). Adventitious reinforcement of maladaptive stimulus control interferes with learning. Behavior Analysis in Practice, 9(3), 223–229. https://doi.org/10.1007/s40617-016-0131-2 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC4999367/)
- Tiger, J. H., Hanley, G. P., & Bruzek, J. (2008). Functional communication training: A review and practical guide. Behavior Analysis in Practice, 1(1), 16–23. https://doi.org/10.1007/BF03391716 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC2846575/)
- Fisher, W. W., Greer, B. D., Fuhrman, A. M., & Querim, A. C. (2015). Using multiple schedules during functional communication training to promote rapid transfer of treatment effects. Journal of Applied Behavior Analysis, 48(4), 713–733. https://doi.org/10.1002/jaba.254 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC4861165/)
- Shamlian, K. D., Fisher, W. W., Steege, M. W., Cavanaugh, B. M., Samour, K., & Querim, A. C. (2016). Evaluation of multiple schedules with naturally occurring and therapist-arranged discriminative stimuli following functional communication training. Journal of Applied Behavior Analysis, 49(2), 228–250. https://doi.org/10.1002/jaba.293 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC4892991/)
- Briggs, A. M., Fisher, W. W., Greer, B. D., & Kimball, R. T. (2018). Prevalence of resurgence of destructive behavior when thinning reinforcement schedules during functional communication training. Journal of Applied Behavior Analysis, 51(3), 620–633. https://doi.org/10.1002/jaba.472 (read at https://pmc.ncbi.nlm.nih.gov/articles/PMC6041172/)