Every option is a real, defensible intervention. Every option is backed by real evidence. And three of them are still wrong.
That's the moment a "best answer" question stops feeling like a knowledge test and starts feeling like a trap. You're not eliminating fake choices anymore. You're ranking four genuinely good ones against each other, under a clock, and picking the single best one is the entire task.
These are consistently the hardest questions on the BCBA exam — not because they're testing obscure content, but because they're testing a skill most study methods never actually practice: weighing multiple real constraints against each other at once, instead of matching one fact to one answer.
Most exam prep trains pattern-matching. See "escape-maintained," think "functional communication training." See "automatic reinforcement," think "competing stimuli." That instinct is correct as far as it goes — it's just not sufficient once a question layers in feasibility, safety, and history on top of function.
A best-answer question doesn't ask "which of these matches the function?" It asks "which of these matches the function and can actually be implemented and accounts for what's already failed and respects every hard constraint in the case?" Multiple filters, applied simultaneously — and most wrong answers only fail one of them. They pass everything else.
On a best-answer question, "this could work" is not the bar. Multiple options will clear that bar. The bar is "this is the option that survives every constraint the stem gave you, not just the ones that come to mind first."
A stem loaded with specific numbers, prior failed attempts, and explicit restrictions isn't scene-setting. Treat those details as potentially diagnostic, not decorative — ask what each one changes about the answer choices.
This one's dense on purpose — real best-answer items usually are. Before you look at the options, pull the constraints out of the stem yourself. That's the actual skill this question is testing, not something to just watch happen in the explanation afterward.
A BCBA is developing a treatment plan for a 7-year-old with autism who engages in severe hand-biting that has caused visible tissue damage. A functional analysis confirms the behavior is maintained by automatic reinforcement. The child has a diagnosed bleeding disorder, and the medical team has explicitly ruled out any procedure involving physical contact or blocking due to elevated injury risk. A competing stimulus assessment has identified specific stimuli that reliably compete with the behavior. An earlier attempt at noncontingent reinforcement — using generally preferred items rather than the validated competing stimuli — failed to reduce the behavior. The classroom team can make the validated competing stimuli continuously available throughout the school day as part of the environment, but only the aide present during one dedicated 20-minute daily 1:1 session is trained to actively redirect the behavior and teach a replacement response. The treatment team has also identified increasing the child's independent, appropriate access to competing stimulation as a treatment goal. Given ALL of these factors, which intervention is MOST appropriate as the primary treatment component?
Keep reading — we're about to weigh all four options against every constraint in the stem, one at a time.
Start with the plain version. The stimuli in option A were actually tested and shown to compete with the hand-biting — that's what the competing stimulus assessment established, so this option is targeting the real function, not just any generally pleasant item. The redirection is explicitly noncontact, so it doesn't run anywhere near the one restriction the medical team stated outright. It also directly answers the team's stated goal — teaching the child an independent way to request access, rather than leaving them dependent on stimuli that happen to already be in reach. And the resourcing matches what's real in this case: the harder-to-deliver piece — active redirection and teaching — lives in the one session where trained staff are present, while the safety-relevant piece — the stimuli themselves being available — doesn't depend on that session at all.
Notice what you never had to decide: whether NCR is a good intervention, whether extinction is a good intervention, or whether DRO is a good intervention. All three are, in the right case. You only had to decide which one survives this case's specific constraints without breaking any of them.
Here's the same comparison laid out directly — which filters each option clears, and which one it breaks.
| Option | Targets the validated function? | No prohibited contact? | Advances the stated independence goal? |
|---|---|---|---|
| A | Yes | Yes | Yes |
| B | Yes | Yes | No |
| C | Yes, in theory | No — requires blocking | Yes |
| D | No — generic preferred item, not the validated stimuli | Yes | No |
Option A is the only row with no red cell. Notice B, C, and D each fail for a genuinely different reason — that's what makes this hard. There's no single shortcut that eliminates all three at once; each one has to be checked against the case on its own terms.
This one uses the correct stimuli — the ones the CSA actually validated, not the generic preferred items that already failed once. That's exactly what makes it tempting: it looks like the direct fix for the prior failure, and a student can reason their way here correctly. It just stops one step short. The team already identified independent, appropriate access to competing stimulation as a treatment goal — and pure response-independent delivery never builds that independence. The child still has no way to get access on their own when the stimuli aren't already in reach.
Mechanistically, this is a textbook answer for automatically maintained behavior — block the sensory payoff, replace it with a rich alternative. If the case had no medical complication, this would be a strong contender. But the stem doesn't leave that door open: blocking access to a sensory consequence means physical contact, and physical contact is exactly what's been explicitly ruled out. A clinically sound mechanism still loses to an explicit, stated restriction.
This one sounds careful — a dedicated assessment, a defined schedule, a familiar procedure. But look at what it's actually built on: a generic preference assessment, not the competing stimulus assessment that specifically validated what competes with this automatically maintained behavior. That's the exact same category error the earlier failed NCR attempt made — a preferred item isn't automatically a competing one. Repeating that mistake with a new procedure name doesn't fix it.
Function-matched using the stimuli that were actually validated, explicitly noncontact, and the only option that teaches the child something to do instead of hand-biting. Nothing borrowed from a different question, nothing assumed the stem didn't state.
A best-answer question isn't asking which option is good. It's asking which option is good and never breaks a single rule the case just gave you.
Most prep programs tell you which answer is correct. BA Prep teaches you why the wrong answers looked right in the first place — including the hardest kind, where every option is individually defensible and only one survives the full weight of the case.
Start Free Trial