Simple discrimination

Consequences establish differential responding

Suppose selecting a green icon produces the requested page while selecting a gray icon does not. Across learning opportunities, selection may become more likely in the presence of green and less likely in the presence of gray.

The icons have discriminative functions because of that history. Their colors alone do not contain behavioral meaning.

SD and S-delta are common labels

An SD signals that a specified response is more likely to contact reinforcement under the current contingency. An S-delta signals that reinforcement is unavailable for that response.

These labels are response and consequence specific. The same green icon can be an SD for opening one page and irrelevant to another response.

Conditional discrimination adds another cue

In conditional discrimination, which stimulus functions as correct depends on another stimulus or context. A sample picture can make one comparison correct, while a different sample makes another correct.

In simple discrimination, that extra conditional relation is absent from the defined arrangement. Calling a three-choice task “simple” based only on the number of stimuli can miss its structure.

Motivating operations answer another question

A discriminative stimulus concerns the availability of a consequence for a response. A motivating operation changes the current value of that consequence and behavior that has produced it.

Both can operate together. A help icon may signal that assistance is available, while a blocked task temporarily increases the value of assistance.

A fictional example

Dev chooses a goal of identifying whether a meeting room is available. A green door display signals that tapping “reserve” produces a booking screen. A red display signals that it does not.

Across ten green opportunities, Dev taps within five seconds in 9 of 10. Across ten red opportunities, tapping occurs in 2 of 10. Report both rates, latency, display order, and any system failures.

This pattern supports differential control by the tested displays. It does not show that color alone controls performance across other devices or contexts.

Balance trials and consequences

Present conditions often enough and in a planned order. Avoid making one stimulus correct far more often or placing it consistently on one side.

Deliver the programmed consequence accurately. A missed consequence following a correct response changes the learning history and should appear in procedural-integrity data.

Analyze errors by stimulus and response

Record which response occurred under which stimulus. “70% correct” can hide perfect responding under one condition and chance-level responding under another.

Measure omissions separately. A person may wait, communicate uncertainty, or decline rather than make the defined selection. Those responses can guide task and access changes.

Generalization needs planned variation

After discrimination is established with trained stimuli, test new examples that preserve the relevant feature while varying irrelevant features. Reserve test items before teaching.

A progressive study of simple and conditional discrimination illustrates one research sequence. It does not create a required order for every learner.

Access can determine apparent performance

Confirm that stimuli are perceptible, response options are usable, and instructions are understandable. Provide needed visual, auditory, tactile, motor, and communication supports.

Do not remove augmentative and alternative communication, glasses, hearing technology, or mobility access to simplify a discrimination task. Document adaptations and system failures.

The BACB outline supplies broad scope

The BACB BCBA Test Content Outline covers stimulus control, discrimination, motivating operations, and procedures for simple and conditional discriminations. It is examination content rather than a client protocol.

A qualified clinician should choose procedures based on assessment, evidence, goals, preferences, context, and scope.

A teaching checklist

• Which response and consequence are defined?
• Which stimuli signal availability and unavailability?
• Is another cue changing which answer is correct?
• Are presentation and position balanced?
• Did programmed consequences occur?
• Are omissions, errors, and system failures separate?
• Which novel examples test generalization?
• Does the skill improve a meaningful daily outcome?

Use a stimulus-by-response matrix

Put each antecedent condition in a row and every observed response in a column. Enter opportunities, response counts, obtained consequences, and median latency. This layout shows whether differentiation reflects more of the target response under the SD, fewer responses under the S-delta, or a shift to another response. Add an implementation column so missed reinforcement or an incorrect prompt is visible beside the affected trial.

Review the matrix across people and settings before claiming generalized control.