Stimulus Discrimination in ABA: 5 Examples & How to Teach It

Two similar instructions, like "touch cup" and "touch plate," should get two different responses. Without stimulus discrimination, a learner may tap whichever object sits closest, so both instructions end up getting the same answer.

In Applied Behavior Analysis (ABA), building that distinction between cues is the goal of discrimination training. Below, you'll find the cue-and-reinforcement mechanics behind stimulus discrimination, five examples from ABA programs, and the teaching and data practices that show when a learner has it.

What is stimulus discrimination in ABA?

Stimulus discrimination in ABA is a learner's ability to change their response based on which of two similar cues is present. Only one of those cues has been followed by reinforcement, a consequence that makes the response more likely to happen again.

In practical terms, a learner with stimulus discrimination notices the specific feature that tells them which response earns reinforcement, such as the shape of an animal's ears or the first sound of a word.

For example, a learner working on receptive labels (pointing to an item when someone names it) might touch any card at first when asked to "touch dog." Once discrimination develops, they touch the dog card when asked for "dog" and the cat card when asked for "cat," even when the cards sit side by side and look similar at a glance.

Psychology textbooks use the same term for classical conditioning, where Pavlov's dogs learned to respond to the one tone that came before food and ignore other tones. ABA programs focus on the operant version, where a learner's own actions change depending on the cue.

How does stimulus discrimination work in ABA?

Stimulus discrimination works through differential reinforcement. One stimulus, the discriminative stimulus (SD), signals that a specific response will be reinforced, while a second stimulus, the S-delta, signals that the same response won't be reinforced.

Over repeated trials, the response becomes more likely with the SD and less likely with the S-delta. A green card on the table might signal that reaching for a toy earns the opportunity to play with it (SD), while a red card signals that reaching won't pay off yet (S-delta).

Out in the community, a learner might walk over when a familiar caregiver calls their name (SD) and stay put when an unfamiliar adult calls it (S-delta).

How fast the learner develops that discrimination often comes down to how different the two stimuli are from each other and how much each one stands out, known in ABA as stimulus disparity and salience.

That's one reason clinicians plan which materials will appear together in early trials, since two look-alike pictures will make the task harder.

What are some examples of stimulus discrimination?

Here's how stimulus discrimination shows up across five common skill areas in ABA programs:

• Language and labeling: A learner says "ball" when shown a ball and "orange" when shown an orange of the same size and shape.
• Social cues: A learner says "hello" to familiar therapists and caregivers, but waits for an introduction before they greet an unfamiliar adult.
• Daily living skills: A learner reaches for a coat on a cold morning and leaves it on the hook on a warm afternoon.
• Academic skills: A learner points to "b" when asked for "b" and to "d" when asked for "d," even though the two letters mirror each other.
• Safety skills: A learner responds differently to a police officer's request to stop than to a stranger's request to come closer.

As the last example shows, a missed discrimination can have grave consequences. In a 2021 study, four autistic children who learned to run from strangers' lures also started running from police officers, though they had followed officers' directions before the training.

The research team corrected that overgeneralization (a response spreading to cues it was never meant for) by adding discrimination training between officers and strangers.

What’s the difference between stimulus discrimination and stimulus generalization?

Stimulus discrimination and stimulus generalization are related but opposite skills that a treatment plan often targets. Discrimination narrows the learner's response to specific stimuli. Stimulus generalization widens the response across new stimuli and settings.

If a learner generalizes a requesting skill, they might ask for "juice" at home, at the clinic, and at a relative's house, using the same word in every setting. Discrimination shows up when that learner asks the adult who usually pours the juice and skips asking a younger sibling who can't reach the fridge.

A treatment plan can call for both skills at once. A learner might need to generalize handwashing to every sink they use while also discriminating between the soap pump and the lotion pump sitting next to it.

How do clinicians teach stimulus discrimination?

Clinicians usually teach stimulus discrimination with three main procedures: errorless learning, prompting, and differential reinforcement.

Errorless learning starts by making the correct response nearly impossible to miss, often through errorless prompting. This involves highlighting the correct stimulus, then gradually prompting less until the learner responds to the natural cue alone.

A 2020 systematic review of 28 studies found that errorless procedures built into the stimuli themselves improved discrimination skills in most cases for individuals with intellectual and developmental disabilities. Most of those studies used stimulus fading, which starts with an exaggerated difference between the choices (such as a faint wrong answer next to a bold right one) and slowly evens out the difference.

As the learner gets more consistent and clinicians fade their prompting, differential reinforcement stays constant: correct responses to the SD get reinforced, and responses to the S-delta don't. That pattern supports discrimination skill-building once the prompts are gone.

How does stimulus discrimination shape treatment planning?

Stimulus discrimination shapes treatment planning by influencing which goals get written, how they're sequenced, and how progress gets measured session to session.

When an assessment shows that a learner mixes up similar-sounding words like "cat" and "cap," a Board Certified Behavior Analyst (BCBA) can write a target for that exact pair and teach it directly.

From there, collected data from SD and S-delta trials shows whether a learner is truly discriminating or getting lucky. With only two cards on the table, a learner choosing at random will still be right about half the time, so the BCBA compares the data against the target's mastery criterion (a set accuracy level across sessions) before determining whether the skill is in place.

Track stimulus discrimination data with Passage Health

Assessing a learner’s progress accurately means reviewing clean SD and S-delta data against a clear mastery bar. Passage Health keeps that session data, your schedule, and your claims in one platform, so nothing has to be re-entered between the session and the progress review.

For teams running discrimination programs, that includes:

• Mobile data collection: Registered Behavior Technicians (RBTs) record their data in the mobile app, and the session note fills in with that data in real time. Upgraded session timers can start frequency, duration, and rate targets automatically in the phases you choose.
• Graphing and treatment reports: Graphs build automatically from the first data point, including a view that shows whether one RBT's trials look different from the rest. Customizable treatment reports pull those graphs into a progress summary you can send for reauthorization.
• Scheduling: Color-coded daily and weekly calendar views show each team member's availability, and drag-and-drop in the daily view makes staff reassignments quick. Team members with scheduling permissions can also create and edit events from the mobile app.
• Billing and utilization reporting: Generate claims from completed sessions, then submit and track them by status, including secondary claims for learners with more than one funding source. The authorization utilization report compares authorized hours with completed and billed hours, and an expiration report sorts authorizations by how soon they end.
• Permissions and signatures: Custom role permissions control who can create, view, and lock session notes. Signature types record who signed each note, in what role, and when, and caregivers can now sign notes from the caregiver portal.
• Client Intake: New referrals move through one intake pipeline, from the first contact to conversion, with caregiver forms sent through the caregiver portal.

Our team also provides one-on-one onboarding, adds new features to the platform every quarter, and offers Frontera AI integration for clinical AI features.

Book a demo to see how your discrimination data, schedules, and claims connect in Passage Health.

Frequently asked questions

How is a discriminative stimulus different from a prompt?

A discriminative stimulus is different from a prompt because the SD is the natural cue that signals reinforcement is available, while a prompt is an extra cue a therapist adds to help the learner respond correctly. As training progresses, the prompt gets faded until the SD alone reliably produces the right response, a state ABA calls stimulus control.

Can stimulus discrimination fade without ongoing practice?

Yes, stimulus discrimination can fade when a learner gets few chances to practice it. In the abduction-prevention study mentioned above, correct responses to strangers and police officers held up unevenly over time, and two of the four children needed booster sessions (short refresher training) to get them back.

Is stimulus discrimination training only used with children?

No, discrimination training applies to learners of any age. In one 2023 study, seven autistic adults improved their job interview skills through remote behavioral skills training (instructions, modeling, and practice with feedback). A teenager or adult can also need the same kind of stimulus-specific teaching as a younger person.

Can a learner discriminate among more than two stimuli at once?

Yes, a discrimination target can include three or more stimuli in the same task. When a learner picks the correct card from an array of five, the other four cards all work as S-deltas for that trial.

Who typically runs stimulus discrimination trials during a session?

An RBT usually runs discrimination trials during a session, working under the close supervision of a BCBA or another qualified supervisor. Since the BCBA judges progress from those trial records, each response needs to be logged as it happens.

References

Arnold-Saritepe, A. M., Phillips, K. J., Taylor, S. A., et al. (2023). Generalization and maintenance. In J. L. Matson (Ed.), Handbook of applied behavior analysis for children with autism: Clinical guide to assessment and treatment (pp. 415–433). Springer. https://doi.org/10.1007/978-3-031-27587-6_21

Behavior Analyst Certification Board. (n.d.). Registered Behavior Technician (RBT). https://www.bacb.com/rbt/

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 

Kahng, S., Butler, C., Kronfli, F. R., et al. (2023). An individualized approach to teaching adults with autism to successfully navigate job interviews via remote instruction. Journal of Applied Behavior Analysis, 56(2), 352–364. https://doi.org/10.1002/jaba.977

Ledbetter-Cho, K., Lang, R., Lee, A., et al. (2021). Teaching children with autism abduction-prevention skills may result in overgeneralization of the target response. Behavior Modification, 45(3), 438–461. https://doi.org/10.1177/0145445519865165 

Markham, V. A., Giles, A. F., Roderique-Davies, G., et al. (2020). Applications of within-stimulus errorless learning methods for teaching discrimination skills to individuals with intellectual and developmental disabilities: A systematic review. Research in Developmental Disabilities, 97, Article 103521. https://doi.org/10.1016/j.ridd.2019.103521 

Spielman, R. M., Jenkins, W. J., & Lovett, M. D. (2020). Psychology 2e. OpenStax. https://openstax.org/books/psychology-2e/pages/6-2-classical-conditioning