The Fading Prompt That Never Faded
A 4-year-old named Priya is learning to request help. The team uses a most-to-least prompting hierarchy: full physical, partial physical, model, gesture, then independent. After eight weeks, Priya requests help independently in 90% of opportunities at home but only 15% at preschool. At school, staff report she rarely encounters tasks she cannot do, and when she does, a teacher aide typically hands her the materials before she asks. The BCaBA has data from home only; school data are collected by the aide on a simple tally sheet, and the aide has had no training on the prompting hierarchy.
1. Is Priya's skill 'mastered'? Justify using the concept of stimulus generalization and response maintenance.
Read the worked answer
No. The 90% figure reflects stimulus generalization failure across settings, not mastery. Mastery requires the response to occur under relevant stimulus conditions - here, at least two settings and more than one person - and to maintain. The home data show the response is in Priya's repertoire and is reinforced there. The school data are uninterpretable for a different reason: the aide's tally sheet has unknown reliability, the aide has not been trained on the hierarchy, and the environment may not even present opportunities. Before concluding anything about generalization, the BCaBA needs to verify that (a) the school actually presents difficult tasks, (b) the aide records opportunities as well as responses, and (c) the recording is accurate. A 15% rate could reflect a true generalization deficit, a measurement artifact, or an absence of opportunities. Only the first would call for programmed generalization; the second and third call for measurement and environment fixes first.
2. Design a generalization plan and identify the key procedural safeguards.
Read the worked answer
The plan should use programming common stimuli and train loosely. Concretely: (1) Identify the discriminative stimuli at school that should occasion help requests (specific task types, the teacher's presence) and incorporate those materials into home or clinic training. (2) Teach across at least two adults and two rooms from the start, rather than training to criterion with one adult and then probing. (3) Vary the tasks, the wording of instructions, and the position of materials so Priya's responding is not tied to incidental cues. (4) Train the aide on the hierarchy and on delayed prompting so she does not preempt the request by handing over materials. (5) Build in natural maintaining contingencies - help must actually produce help at school, promptly and consistently. (6) Use recruitment of natural communities: the classroom teacher, not just the aide, should respond to requests. Safeguards: do not remove the home supports until school data are stable; collect generalization probes before, during, and after training; and ensure the aide's data include opportunities, responses, and prompt levels so the BCaBA can separate a skill deficit from a measurement problem.
3. The aide asks whether she should stop helping Priya entirely so Priya 'has to ask.' How should the BCaBA respond?
Read the worked answer
Decline that plan. Withdrawing all help would create a motivating operation for escape and possibly for problem behavior, and it risks punishing approximations of the request. The aide's help is the reinforcer for the mand; removing it removes the consequence that maintains the skill. The better arrangement is contingent help with a brief delay: the aide pauses a few seconds to give Priya an opportunity to request, prompts at the least intrusive level needed, and then provides help immediately and enthusiastically when Priya asks. Over sessions the delay lengthens and prompts fade. The BCaBA should also confirm that the tasks used are genuinely difficult enough to occasion help-seeking - if Priya can complete everything, there is no EO for the mand and the low rate is expected, not a deficit. This distinction (skill deficit vs. performance deficit vs. absent EO) is the core judgement call and should be documented in the plan.