The Plan That Doesn't Match the Data
Priya, 6, attends a clinic three afternoons a week. Over 12 days an RBT collected ABC data on aggression: 47 incidents. Thirty-one followed removal of a plastic ball or denial of the sensory swing, 6 occurred during transitions between rooms, 4 followed the end of a preferred video, and 6 had no clear antecedent recorded. The BCBA left for a two-week vacation and wrote in the plan: 'Function = escape from demands. Start escape extinction Monday, no exceptions.' On Monday the classroom aide tells the RBT to go ahead and run a demand condition alone today, since 'the BCBA already figured out the function.'
1. Do the ABC data support the conclusion that the function is escape from demands?
Read the worked answer
No - and the mismatch is large, not a matter of interpretation.
- 31 of 47 incidents (about two-thirds) followed loss of a tangible item or denied access, not a demand.
- 6 were transitions, 4 were termination of a video (also loss of a tangible/activity), 6 unaccounted.
- Zero incidents are described as following a demand.
The pattern points toward access to tangibles/activities. The stated function in the plan contradicts the data the plan is supposedly based on.
A second issue: ABC data are descriptive and correlational. They generate a hypothesis; they do not demonstrate a functional relation, because no condition was manipulated and no comparison was made. Observer attention also biases what gets recorded - the six 'no antecedent' incidents are a reminder of that.
The RBT's obligation is to flag the discrepancy to whoever is covering supervision before the plan is run, not to execute a plan whose stated function the data contradict. Implementing a contingency built on the wrong function risks strengthening the very behavior targeted.
2. The aide wants the RBT to run a demand condition alone. What should the RBT do?
Read the worked answer
She should decline to run it alone and escalate, politely and immediately.
Reasons:
- Scope: RBTs implement assessment and intervention procedures under a supervising behavior analyst's direction. They do not design functional analyses or independently conduct experimental conditions. 'The BCBA already figured out the function' is not supervision for a procedure the BCBA did not write for that day.
- Clinical risk: a demand condition deliberately withholds escape. If the actual function is tangible, the condition tells you nothing, and Priya spends a session in escalation for no scientific gain.
- Safety: aggression is the target; without a supervisor present and without written termination criteria, the RBT and aide are managing a possible burst on their own.
What she can do instead: contact the covering supervisor or clinical director the same day, send the ABC summary with the count breakdown, ask for written direction, and document the contact attempt. Until then, continue the existing structured data collection and the current (non-extinction) procedures - not a new condition invented in the hallway.
3. Staying inside her role, what could the RBT reasonably propose to fill the two weeks productively?
Read the worked answer
Several things are within scope when the supervisor authorizes or has already written them:
- Tighten the descriptive data. Add times, activity, who was present, and a clear antecedent code list so the six unrecorded antecedents stop happening. Consider a scatterplot by time of day to see whether incidents cluster.
- Run a structured preference assessment the BCBA has already specified in the plan - a brief MSWO or a free-operant observation with the candidate items (ball, swing, videos) - to identify what actually functions as a reinforcer. This is data the team will need whichever function wins.
- Document what already works: which consequences currently de-escalate Priya, and how staff respond now (baseline treatment integrity).
- Prepare a request to the supervisor listing the discrepancy and asking for a decision on record.
What she should not do is independently choose a function and start an extinction contingency on it. Gathering information is in scope; drawing the clinical conclusion and changing the plan is not.