Comparing Methods for Evaluating Sensitivity to Establishing Operations and Bias Toward Challenging Behavior
When assessing challenging behavior, the way you set up consequences can shape what you observe—and what you might miss. This study examines whether reinforcing both appropriate and challenging behavior during a brief functional analysis yields different clinical information than reinforcing only severe challenging behavior. The findings have practical implications for how clinicians plan prevention strategies and gauge escalation risk.
What is the research question being asked and why does it matter?
The researchers asked a straightforward practice question: when you run a brief, trial-based “mini functional analysis” that tracks both appropriate and challenging behavior, should you reinforce both the same way—or only reinforce severe challenging behavior?
This matters because the way you arrange consequences during assessment can change what you see. If what you see changes, your “risk read” can change too, affecting what you choose to teach and prevent.
The study focused on two ideas useful for prevention planning. The first is sensitivity to common establishing operations (EOs), like losing attention, facing hard work, or losing toys. Sensitivity here means: does this situation reliably bring out problem or request behavior?
The second idea is bias—whether the child tends toward more challenging behavior than appropriate behavior when both could work.
If an assessment setup hides appropriate behavior (because it’s not reinforced), you might conclude a child “doesn’t request” when they actually would if it worked. If a setup hides severe behavior (because requests quickly end trials), you might miss that a child can escalate. Clinicians need both kinds of information to plan safe, respectful prevention and teaching.
What did the researchers do to answer that question?
They worked with six preschool-aged children (3–5 years old) in a Head Start program. These were neurotypical children referred by teachers for concerns about challenging behavior. Sessions took place in a small room near the classroom, using a play-based format with common classroom-like events.
They ran repeated “BASE” sessions, each containing 15 short trials. Every trial had a 1-minute control segment (access to toys, attention, and no demands) followed by a 1-minute test segment that created one EO. The three EOs were attention (adult withdraws), escape (adult gives instructions), and tangible (adult removes toys).
They compared two consequence arrangements. In the symmetrical setup, the first instance of either appropriate or challenging behavior (Level 1 or higher) ended the test and produced the matching reinforcer. In the asymmetrical setup, only the most severe challenging behavior (Level 4) produced the reinforcer. Appropriate behavior received a brief “thanks for asking, not right now” response while the EO stayed in place.
Behavior was recorded using a Neutral-to-Severe rating scale capturing both appropriate behavior (simple to complex requests) and challenging behavior (mild to severe, including unsafe forms). Rather than simply counting behavior, they tracked the highest level shown in each segment. They then compared whether test segments produced more behavior than control segments (sensitivity) and whether test segments showed more challenging than appropriate behavior (bias).
How you can use this in your day-to-day clinical practice
If you want a clearer read on “which EOs matter,” start by reinforcing both appropriate and challenging behavior during assessment. In this study, symmetrical contingencies more often showed sensitivity—more behavior in test than control—especially for tangibles.
Practically, if you’re mapping which situations trigger behavior, a setup that lets requests “work” may reveal the actual reinforcers faster and with less extinction noise. Use that information to prioritize what to teach first.
If your main concern is “how bad can this get if the child doesn’t get what they want,” an asymmetrical setup may show more severe behavior. The asymmetrical arrangement in this study captured more severe challenging behavior overall.
Clinically, if you’re building a safety plan, staffing plan, or gauging treatment urgency, you may need an arrangement that doesn’t immediately end the EO when a child uses a low-level response. Otherwise, you might underestimate escalation risk. This should only be done with strong safety precautions, clear stop rules, and only when the potential benefit outweighs the risk of evoking severe behavior.
Plan to use the two approaches for different decisions, not as competitors. Use symmetrical trials when the goal is prevention planning and identifying what to reinforce—it avoids repeatedly putting appropriate behavior on extinction. Then, if needed for safety planning, carefully add limited asymmetrical trials to see whether the learner escalates when appropriate behavior doesn’t work right away.
Don’t assume you must always run the asymmetrical version. This study’s participants were young and referred for concerns; results may differ for learners with different histories, diagnoses, or strong communication skills.
When you interpret “low requesting” during assessment, check the contingency before concluding there’s a skill deficit. In the asymmetrical setup, appropriate behavior didn’t access reinforcement, which can suppress responding even when the child is capable.
If you see few requests during an asymmetrical-style format, treat that as “requests didn’t work here”—not automatically “the child can’t request.” For treatment planning, still assess the learner’s communication repertoire through direct teaching probes, natural environment data, and caregiver report before setting goals.
When you see sensitivity without strong appropriate behavior, treat that as a teaching signal. If toy removal produces lots of behavior but almost no functional requesting, that’s a practical risk marker: the EO is common in real life, and the child may lack a reliable, dignified way to get needs met.
Build simple, fast, high-probability communication responses tied to that EO. Arrange the environment so those responses contact reinforcement often at first, then thin based on data rather than starting with denial.
Don’t use “bias toward challenging behavior” as a label for the child. Use it to describe the current choice pattern under specific conditions. In this study, bias toward challenging behavior showed up under both setups for most children, especially during tangible trials.
This suggests you shouldn’t assume reinforcing requests during assessment will automatically eliminate challenging behavior. Plan for both: teach the request, and reduce payoff for unsafe behavior while increasing payoff for safe behavior, using the least intrusive approach that keeps everyone safe.
Be cautious about generalizing these findings. These were six neurotypical preschoolers in a specific setting, and the order wasn’t fully counterbalanced (symmetrical was always first). Extinction exposure in the asymmetrical condition could have carryover effects. The study also didn’t test interventions afterward, so you can’t assume a particular assessment format leads to better treatment outcomes.
What you can do is choose assessment contingencies on purpose, based on the exact clinical question you’re trying to answer—while prioritizing safety, assent, and minimizing unnecessary evocation of severe behavior.
Works Cited
Allen, A. E., Bridges, K. G., Pizarro, E. M., & Morris, S. L. (2026). Comparing methods of evaluating sensitivity to common establishing operations and bias toward challenging behavior. *Journal of Applied Behavior Analysis, 59*(1), e70046. https://doi.org/10.1002/jaba.70046
