In May 2026, Ashlyn McChristie and David A. Wilder published Teaching Young Children With Autism: When Not to Cooperate With Instructions in the Journal of Positive Behavior Interventions. Wilder is Professor and Head of School of Behavior Analysis at Florida Institute of Technology. The article appears during a period of increasing criticism directed at compliance-centered autism intervention, particularly criticism surrounding ABA’s historical emphasis on obedience and behavioral conformity.

The researchers position the study as a corrective. Rather than teaching autistic children to comply indiscriminately with adult instructions, the article proposes “discriminated cooperation,” meaning children should learn when cooperation is appropriate and when refusal is appropriate. The project frames this distinction as an abuse-prevention and self-advocacy skill. Allegedly, “teaching young children to discriminate when they should and should not cooperate with adult instructions may prevent sexual and perhaps other forms of abuse.”

Three autistic children participated. Mary was a 5-year-old white girl receiving approximately 20 hours per week of in-home behavioral services. Joe was a 4-year-old white boy receiving approximately 30 hours weekly in a clinic setting. Richard was a 6-year-old white boy also receiving approximately 30 hours per week in a clinic environment. All three children were described as having mild intellectual disabilities and relatively strong receptive language skills. The article specifically identifies all three children as “white,” although race plays no role in the design, interpretation, or findings. The descriptor functions as demographic bookkeeping rather than meaningful social context.

The Clinic as Experimental Space

The study took place entirely inside institutional treatment environments already organized around adult instruction and behavioral observation. Mary’s sessions occurred in a university clinic room equipped with a one-way mirror. Joe and Richard participated in clinic rooms monitored through video cameras because no observation mirror was available.

The experimenter was a woman in her mid-20s who had previously worked as a therapist in the children’s clinic. She was therefore already familiar to the participants before the experiment began. The study was not occurring between strangers. It was occurring between children and an adult already associated with treatment, instruction, and behavioral authority.

The room itself functioned as a tightly controlled behavioral environment. The researchers controlled who entered the room, where the adult stood, when instructions were delivered, what toys were available, and what consequences followed each response. The setting resembled a laboratory of compliance more than a natural social interaction. Every element of the child’s environment was organized around measurement.

What the Children Were Told to Do

The experiment centered on a single repeated instruction: “Take your shirt off.” Sometimes the adult delivered the instruction while standing outside the room. Sometimes she delivered the exact same instruction while standing inside the room with the child. One condition was labeled “should cooperate.” The other was labeled “should not cooperate.”

Each session involved ten separate trials. Five times the child was expected to obey the instruction and remove their shirt. Five times the child was expected to refuse the instruction and keep their shirt on. The order was randomized. Cooperation was defined as removing the shirt within 10 seconds of instruction delivery. Refusal meant not removing the shirt.

If the child removed their shirt, the researchers waited approximately three minutes before placing the shirt back onto the child “without vocal interaction.” In plain language, the children sat shirtless inside therapy rooms while researchers intentionally minimized conversation because emotional interaction itself could interfere with the behavioral data. The child’s comfort, embarrassment, confusion, or interpretation of the interaction became secondary to procedural consistency.

Rewards, Reinforcement, and Grooming Logic

The intervention phase used Behavioral Skills Training combined with video modeling. The children first received verbal rules about privacy. They were told that removing their shirt was acceptable only when alone in the room and unacceptable when another person remained present. They then watched videos of another child correctly following these rules.

After watching the videos, the participants rehearsed the responses themselves. Correct responses received praise and “highly preferred tangible items.” The article never specifies exactly what those rewards were. In ABA research, “tangible items” commonly means favorite toys, snacks, electronics, stickers, or other objects identified beforehand through preference assessments. The children therefore learned through behavioral reinforcement. Certain bodily responses under certain environmental conditions produced immediate approval and access to desired objects.

The study frames this procedure as abuse prevention. The researchers argue that autistic children may be vulnerable to abuse because disabled children experience higher abuse rates and because many autistic children are heavily trained to comply with adult instructions. The article specifically states that instructions involving undressing “might be used by child abusers as part of the abuse grooming process.” The shirt-removal exercise therefore becomes a simplified behavioral analog for sexual grooming dynamics.

The Limits of Behavioral Self-Advocacy

The researchers later acknowledge that the scenario was only “a simplified analog rather than a direct model of real-world safety situations.” This admission reveals the central limitation of the study itself. Real grooming and coercion rarely operate through a single environmental cue like whether another adult is physically present in a room. Abuse dynamics often involve secrecy, trust, dependency, manipulation, authority pressure, gradual boundary violations, and familiar adults.

The study reduces this complexity into a binary discrimination task: adult present versus adult absent. The child is not taught how to recognize coercion relationally, communicate discomfort, evaluate trust, or seek help socially. The child is taught a rehearsed environmental rule tied to one observable stimulus. Refusal itself becomes another operant response category shaped through reinforcement, modeling, and repetition.

Success was defined procedurally. Participants had to achieve “three consecutive sessions with 100% accurate responding in both the should and should not cooperate conditions.” The language throughout the paper emphasizes fidelity, procedural integrity, mastery criteria, and accurate responding. These terms measure procedural consistency rather than human understanding. The autistic children’s participation demonstrates that the method can be successfully installed under tightly controlled institutional conditions. Their role is to confirm portability, not lived meaning. This study exists to preserve behaviorism’s authority under increasing ethical criticism while ensuring that the procedural foundations of compliance training remain intact. It demonstrates continuity, not care.

Misbehaviorism Critique

The study is not simply an isolated experiment conducted by one young person with a bachelors’ degree. This misbehaviorism was a supervised production occurring inside a behavior analysis training program, a clinical service environment, a peer-review system run by behavior analysts, and a publication pipeline governed by behavior-analytic journals. The mentorship structure guarantees fidelity to method, not expansion of perspective.

This matters because studies like this are later presented to parents as proof that ABA is “evidence-based,” “scientifically validated,” and “effective.” Parents are rarely shown what that evidence actually looks like in practice. The phrase evidence base evokes images of broad developmental outcomes, emotional well-being, independence, communication, or long-term quality of life. What many of these experiments actually measure is far narrower: whether a child performs a specified behavior under tightly controlled institutional conditions after repeated prompting, rehearsal, and reinforcement.

In this case, the experiment involved three autistic children inside therapy rooms being repeatedly instructed by a familiar adult to remove their shirts under alternating environmental conditions. Correct responses were shaped through praise, tangible rewards, repetition, video rehearsal, and behavioral contingencies until the children achieved “100% accurate responding.” This is what procedural success looks like inside the ABA evidence base.

The article treats this outcome as evidence that autistic children can learn abuse-prevention skills. The actual measured outcome is narrower. The study demonstrates that children already immersed in intensive behavioral environments can learn to discriminate between two rehearsed stimulus conditions tied to one repeated instruction. The distinction matters because “evidence-based” often functions rhetorically rather than descriptively in conversations with parents. The institutional prestige of science obscures the extremely specific and artificial nature of many of the underlying experiments.

The structure of the research itself reinforces this narrowing process. Graduate trainees implement tightly scripted procedures under faculty supervision. Behavioral journals evaluate whether procedural fidelity, reinforcement schedules, operational definitions, and mastery criteria conform to disciplinary expectations. Success means the procedure worked as designed. The system reproduces itself by training researchers to generate additional procedurally successful demonstrations of behavioral control.

None of this means parents are wrong for seeking help or support for their children. The issue is transparency. When families hear that ABA is supported by “decades of evidence,” they are rarely invited to examine the actual texture of that evidence: children sitting shirtless in clinic rooms while adults withhold interaction to protect internal validity; preferred objects used to shape bodily responses; social meaning reduced into measurable contingencies; and human vulnerability translated into stimulus discrimination tasks.

---

APA Citation: McChristie, A., & Wilder, D. A. Teaching Young Children With Autism: When Not to Cooperate With Instructions. Journal of Positive Behavior Interventions, 10983007261437727.