Skill acquisition & teaching Glossary

Start with the skill and the learner

Define the outcome in observable language and ask whether it matters to the person. Identify ordinary supports, communication modes, motor demands, materials, settings, partners, health needs, and the ways the learner can accept, pause, or decline instruction.

The BACB BCBA Test Content Outline, 6th edition includes prompting, prompt fading, chaining, trial-based and free-operant procedures, observable goals, contextual fit, integrity, and evaluation. It is examination content rather than a universal teaching protocol.

The current BACB Ethics Code addresses competence, client involvement, consent and assent when applicable, medical needs, assessment, risk, data, and continuing evaluation for covered behavior analysts. Communication, AAC, bathroom access, food, water, mobility, prescribed care, and emergency help should not depend on lesson performance.

Chaining organizes multi-step tasks

A task analysis divides a practical activity into observable steps. Forward chaining teaches the first unmastered step, with the instructor or another planned method completing later steps. Backward chaining teaches the last unmastered step first while earlier steps are completed through the selected support.

Total-task presentation gives the learner an opportunity to participate across every step during each teaching episode, with prompts provided as needed. The best choice can depend on skill length, safety, motor effort, current repertoire, consequence contact, preference, and the feasibility of completing untrained steps.

Bancroft and colleagues compared several completion procedures within forward-chaining instruction for seven autistic children. All produced learning under the study conditions, with efficiency differences across participants and measures. This small study supports individualized comparison rather than one standard completion rule.

Prompting needs an exit plan

Prompting adds assistance intended to increase the chance of a correct or successful response. Define the exact prompt, who provides it, when it occurs, how the learner experiences it, and what counts as independent.

Least-to-most prompting begins with less assistance and adds support when the response does not occur or is incorrect. Most-to-least prompting begins with stronger assistance intended to prevent or reduce errors and systematically fades it.

Time delay inserts or increases a planned interval between the natural cue and an added prompt so the learner has a chance to respond. Stimulus fading gradually changes an extra stimulus feature, such as size, contrast, position, or emphasis, until the intended cue controls responding.

Prompt order, delay, response window, error rule, fading criterion, and mastery rule are separate decisions. Watch for prompt dependence, distress, position bias, inaccessible cues, slow partner response, and performance that disappears outside teaching.

Teaching formats serve different purposes

Discrete-trial teaching arranges repeated, clearly bounded teaching opportunities with an antecedent, learner response, consequence, and brief interval. The CDC autism treatment page describes DTT as a stepwise ABA teaching style. That public overview does not specify an individual program.

Natural-environment teaching arranges instruction within activities, routines, materials, and motivations that occur in everyday contexts. Naturalistic does not mean unplanned. The target, opportunity, prompt, partner response, data rule, and generalization plan still need definition.

Structured trials can create many controlled opportunities. Natural contexts can improve relevance and contact with everyday cues. A program may use both while tracking whether learning transfers beyond the teaching arrangement.

Error correction is instruction after an error

Error correction is a planned response after an incorrect, incomplete, or absent response. It may include feedback, a prompt, practice, a transfer trial, or another teaching step. Define what happens, how long it takes, and whether reinforcement follows corrected performance.

Leaf and colleagues compared error-correction and errorless-learning procedures with 28 autistic participants on a narrow tact-learning task. Both produced learning, while response patterns differed. The result does not establish a universal winner, and learner experience should be measured alongside acquisition.

Separate teaching from proof of learning

Report at least four things separately:

• correct or successful responses during teaching
• prompt level and implementation integrity
• performance on planned independent probes
• maintenance and generalization across time, people, materials, and settings

Suppose Amari is learning a five-step lunch-packing routine. During eight teaching opportunities, Amari completes 31 of 40 steps with the planned prompts. On four later unprompted probes, 13 of 20 steps are independent. Teaching success is 31/40, 77.5%; independent probe performance is 13/20, 65%. Combining them would hide the difference.

The team asks Amari which supports help, checks that AAC and visual information are available, and samples the routine at home and school. A skill should remain useful with ordinary supports rather than look independent only in a clinic drill.

Record Amari's effort, preference, errors, pauses, and requests for help alongside step accuracy. Those observations can reveal when the teaching arrangement needs revision.

The ABAI basic-principles page lists broad higher-education content. It is not an implementation manual or endorsement of a case plan.

Explore clinical roles at Finni practices. Ask how clinicians select meaningful goals, compare teaching options, protect assent and communication, fade prompts, and measure generalization.