ABAT Generalization and Maintenance Strategies Flashcards

A behavior technician has successfully taught a child to request a break using a picture card in the clinic. However, the child's teacher reports the child never uses this skill at school. The MOST likely reason for this failure to generalize, and the BEST corrective strategy, is:

Answer: The discriminative stimuli controlling the behavior (specific picture card, specific room, specific therapist) were too narrow; program common stimuli by incorporating natural school materials and personnel into training

Failure to generalize across settings is classically explained by stimulus overselectivity — the child's response came under control of stimuli specific to the clinic (the particular card design, room, therapist). Programming common stimuli involves deliberately bringing natural stimuli from the generalization setting (school materials, school personnel) into training so the controlling variables overlap. While staff training and preference assessment are valuable adjuncts, they do not directly address the root stimulus-control problem.

During maintenance probes 6 weeks after mastery, a client's previously acquired tacting repertoire has dropped from 95% to 60% correct. The technician has been following the thinning schedule as written. Which explanation BEST accounts for this pattern, and which strategy is MOST appropriate?

Answer: The reinforcement schedule was thinned too rapidly relative to the behavior's resistance to extinction; return to the last effective schedule density and thin more gradually using a leaning criterion

A drop in maintenance performance following schedule thinning most parsimoniously indicates the schedule was thinned faster than the behavior's resistance to extinction could support. The evidence-based correction is to return to the last schedule density at which behavior was stable and then thin more gradually, using a performance criterion (e.g., ≥90% across 3 sessions) before each thinning step. Resurgence, probe reactivity, and motivating operations are secondary considerations not supported by the described pattern alone.

A practitioner wants to promote generalization of a social greeting skill to novel peers. She trains with three peers and then tests with untrained peers. Responding does NOT generalize. She switches her approach and trains sufficient exemplars until generalization IS observed. What is the MOST precise technical explanation for why the second approach succeeded?

Answer: Repeated exposure to varied exemplars allowed the client to extract the common relational features of the stimulus class, establishing a generalized operant

Training sufficient exemplars works because exposure to a range of varied instances allows the learner to contact the abstract defining features of the stimulus class (in this case, 'approaching person who could be greeted') rather than the idiosyncratic features of individual training peers. This establishes a generalized operant — a functional response class controlled by the common relational properties of the class. This is distinct from mere fluency, reinforcement history accumulation, or anxiety reduction.

A BCBA is designing a generalization plan for a teenager learning to manage anger using a self-management checklist. She wants the skill to persist without external monitoring. Which combination of strategies MOST directly addresses both generalization and long-term maintenance with the LEAST ongoing therapist involvement?

Answer: Program multiple exemplars of anger-provoking situations during training, then transfer stimulus control to the client's self-monitoring checklist and shift to natural community reinforcers

The most complete strategy for promoting independence combines: (1) multiple exemplars to address generalization across situations, (2) transfer of stimulus control to self-management tools (checklist) so external prompts fade, and (3) contact with natural reinforcers (real-life positive social consequences of controlled behavior) to sustain the behavior without therapist-delivered consequences. Weekly boosters and caregiver-delivered reinforcement increase dependence on programmed contingencies rather than natural ones, counteracting true maintenance.

A behavior technician notices that a client who mastered a chain of steps for preparing a sandwich in training always skips step 5 (spreading condiment) during school lunch, even though all other steps generalize perfectly. The MOST likely behavioral explanation for this step-specific failure is:

Answer: The stimulus that serves as the SD for step 5 in the natural environment differs from the training stimulus, and no generalization training targeted that transitional cue

In behavioral chains, each step is brought under the control of a specific SD — usually the product of the previous step. If the stimulus that sets the occasion for step 5 in training (e.g., a specific condiment container, placement, visual cue) differs meaningfully from the school lunch environment, the link between the step-4 product and the step-5 behavior will not transfer. This is a stimulus generalization failure specific to that chain link. While re-training to criterion and MO analysis are useful considerations, the pattern of a single chain step failing while others generalize specifically implicates a transitional-cue mismatch.

A practitioner uses 'train loosely' as a generalization strategy by randomly varying non-essential elements of training (location within a room, therapist's clothing, time of day, tone of voice). A colleague argues this approach risks undermining acquisition. Under which condition is the colleague's concern MOST technically justified?

Answer: When the target behavior has not yet reached a stable mastery criterion, because loose training during early acquisition may prevent a clear SD-response relationship from forming

Train loosely deliberately introduces variability in non-critical stimulus dimensions to prevent overly narrow stimulus control. However, this strategy is contraindicated during early acquisition before the basic SD-response-reinforcer relationship is established, because premature variability can impede discrimination learning and slow the formation of a reliable operant. The colleague's concern is most justified here: loose training is optimally applied after acquisition is stable, not before. The other conditions described do not represent technically valid reasons to avoid this strategy.