Most of what your clients do well depends on doing it at the right time and place. Stimulus control is the name for that fit between behavior and its surroundings, and recognizing it (or its absence) is a core analysis skill.
Why it matters
A skill that happens at random is not much use. A child who says "bathroom" constantly, or never, has not learned to request the bathroom; a child who says it when they need to go and an adult is nearby has. The same applies to staff: a technician who delivers praise only while you are watching shows that your presence, and not only the client's behavior, is controlling their praise.
When a case describes a pattern of responding and asks what the antecedent is doing, look for a difference in responding across stimulus conditions and trace it back to a history of differential consequences.
Key ideas
- Stimulus control is a relation, shown by differential responding. A response is under stimulus control when some feature of it (rate, latency, duration, or magnitude) reliably differs in the presence versus the absence of a particular antecedent stimulus.
- It is learned through differential reinforcement. When a response is reinforced in the presence of one stimulus and not reinforced (extinction) in the presence of another, responding gradually becomes more likely in the first condition and less likely in the second.
- Discriminative stimuli evoke; they do not elicit. An SD makes a response more likely because of a history of reinforcement in its presence. It does not force the response the way a reflexive stimulus elicits a reflex.
- S-deltas matter just as much. The S-delta (SΔ) signals that the response has not paid off there. A clear contrast between SD and S-delta is what sharpens control.
- You need evidence, not history alone. Reinforcing a response in the presence of a stimulus is the procedure; differential responding is the outcome. Until you see the difference in the data, you cannot claim stimulus control.
- Control can attach to the wrong stimulus. Learners respond to whatever reliably predicts reinforcement, which is sometimes a position, a teacher's glance, or the order of trials rather than the stimulus you intended.
Definitions
- Stimulus control: a change in some dimension of a response that occurs reliably in the presence of a particular antecedent stimulus, compared with its absence.
- Discriminative stimulus (SD): an antecedent stimulus in whose presence a response has been reinforced, so the response is more likely when it is present.
- S-delta (SΔ): an antecedent stimulus in whose presence a response has not been reinforced, so the response is less likely when it is present.
- Differential reinforcement (for discrimination): reinforcing a response in the presence of one stimulus while withholding reinforcement for that response in the presence of other stimuli.
- Faulty stimulus control: control of a response by an irrelevant or unintended stimulus feature rather than the one the teacher meant to teach.
Examples from a session
Reading stimulus control in everyday settings
- Clinic, early learner: When a therapist holds up a cup and asks "What is it?", Mateo, 4, says "cup" within two seconds. When she holds up a spoon and asks the same question, he says "spoon," not "cup." The cup (with the question) is functioning as an SD for saying "cup."
- Home, teen: Rosa, 15, asks to use the car mostly when her mom is around rather than her dad, because her mom usually says yes and her dad usually says no. Mom's presence is an SD for asking; dad's presence is an S-delta.
- School, adult staff: A paraprofessional runs the token system carefully when the BCBA is in the classroom and loosely when she is not. The BCBA's presence exerts stimulus control over the staff member's implementation.
- Faulty control: Jae, 6, touches the correct picture during matching, but only when it is placed on the right side. When placement is randomized, his accuracy falls to chance. Position, not the picture, has control.
A useful habit: whenever you claim a stimulus controls behavior, ask "compared with what?" and name the condition in which the response is less likely.
Common mistakes / exam traps
- Equating high accuracy with stimulus control. If the learner responds the same way whether or not the stimulus is present, there is no evidence of control, however often the response occurs.
- Calling an SD an eliciting stimulus. Elicitation is respondent; an SD evokes operant behavior because of past consequences in its presence.
- Missing faulty control. If performance collapses when an incidental feature (position, order, the teacher's look) changes, suspect the learner was responding to that feature all along.
S-delta vs. discriminative stimulus for punishment (SDp)
Both lower responding, so ask what happened to the response in that stimulus's presence. Reinforcement was withheld (extinction): S-delta. The response was punished, even while reinforcement stayed available: SDp. A stimulus that signals a reprimand is not an S-delta.
SD vs. motivating operation
Ask whether the stimulus signals that the reinforcer is available or changes how much it is worth. Availability: SD (the vending machine's lit display signals that buying works). Value: motivating operation (skipping lunch makes a snack more effective). One case can contain both.
Quick check
Nia, 8, says “hi” on 9 of 10 occasions when a classmate greets her. On 9 of 10 occasions when no one has greeted her, she also says “hi” as a classmate walks by. Is her “hi” under the stimulus control of classmates' greetings?