The ABA debate — what autistic adults say, what the research shows

Applied Behaviour Analysis is the most-debated approach in autism support, and the debate is genuinely live — not a marketing dispute, not a fringe issue. Some autistic adults describe ABA as the most harmful thing that happened to them in childhood. Some clinicians point to outcome studies as evidence of meaningful gains. Both groups are often arguing about different versions of "ABA" without realising it. This piece tries to trace the history, lay out the critique, and summarise what the research actually shows.

This is one of the topics where Soira deliberately does not take a clinical position. We do take the editorial position that the autistic-adult perspective deserves serious weight in any decision a family makes.

Lovaas and the origins Modern ABA traces to Ivar Lovaas at UCLA in the 1960s and '70s. Lovaas's autism work used behavioural conditioning to shape children's behaviour toward what was then considered "normal." The goal of the early UCLA programme was stated explicitly: to make autistic children "indistinguishable from their peers."

Two things about Lovaas's original work are important to know:

It used aversives. Early Lovaas-protocol ABA used slaps, electric shocks, and other physical punishments for "incorrect" responses. This is not a misrepresentation; it is documented in Lovaas's own published work and in interviews. Aversives were standard early-ABA practice into the 1980s.

Lovaas also conducted gay-conversion research. Using similar behavioural-conditioning techniques, the UCLA group ran programmes intended to make "sissy boys" more masculine. The same lead author, the same lab, the same time period, the same theoretical framework. This historical link matters for the autistic-adult community because it underlines that the original goal was conformity, not flourishing.

The 1987 Lovaas paper claimed strong outcomes from intensive (40 hours per week) early ABA. The methodology has been heavily critiqued since — selection bias, control-group issues, outcome measures focused on appearing typical rather than thriving — but the paper remains the foundation document of modern ABA's clinical claims.

What modern ABA looks like Aversives are no longer mainstream in licensed ABA practice. Most contemporary ABA uses positive reinforcement, often in naturalistic settings, and many providers describe themselves as "neurodiversity-affirming." Some genuinely are; some use the label without the substance.

Modern ABA is highly variable. The same credential — BCBA (Board Certified Behaviour Analyst) — covers practitioners running 40-hour-a-week tabletop programmes targeting compliance and practitioners running gentle, child-led, parent-coaching models that look closer to a developmental approach than to classical ABA. The variance between providers under the same name is enormous.

Common features of contemporary ABA:

• Behavioural goals — defined target behaviours to increase or decrease, measured systematically.
• Reinforcement — usually positive, sometimes ignored ("planned ignoring") rather than punished.
• Data collection — sessions are logged, outcomes are tracked, progress is reviewed regularly.
• Manualised curricula — some programmes (Lovaas-protocol, Skinnerian DTT) use structured drills; others (PRT, ESDM-derived ABA) blend developmental and behavioural methods.

The autistic-adult critique The most influential critique has come from autistic adults who were themselves enrolled in ABA as children. The core concerns:

Compliance as the goal. Older ABA's success metric was an autistic child who behaved more like a non-autistic one. Critics argue that even modern ABA often retains this orientation — its goal is the child fitting in, rather than the child flourishing as themselves. Stimming is often targeted for reduction; "appropriate" eye contact is often a goal; "indistinguishable from peers" is sometimes still in protocols.

Suppression of autistic communication. Many autistic adults describe being trained to suppress communicative behaviours — scripting, echolalia, hand-flapping — that they later realised were how they communicated. Trained out of natural communication, they often ended up with less, not more.

The Kupferstein study. A 2018 study by Henny Kupferstein in *Advances in Autism* surveyed adults exposed to ABA in childhood and reported elevated PTSD symptom levels compared with non-exposed autistic controls. The study has methodological limits (self-selected sample, retrospective recall) but it crystallised concerns that had been circulating in the autistic-adult community for years.

Power asymmetry. A young child cannot meaningfully consent to a 40-hour-a-week intervention. ABA's traditional intensity means a level of adult control over a child's life that critics argue is incompatible with the child's developing autonomy.

Position of autistic-led organisations. The Autistic Self Advocacy Network and most autistic-led organisations are opposed to ABA in its current dominant forms. This is not a fringe position within the autistic community — it is the majority view.

The research base Outcome studies on ABA are mixed and methodologically contested. A short summary:

Lovaas's claims of near-recovery rates have not replicated in independent studies at the same effect size.

Modern intensive behavioural intervention (EIBI) shows meaningful gains in IQ scores, adaptive behaviour, and early language in some studies — though gains often plateau, and the outcomes measured rarely include the child's quality of life or the autistic person's adult assessment of the experience.

The 2020 Cochrane review on EIBI for autism concluded that evidence quality remains low to moderate, with significant risk of bias across most studies. Cochrane is the gold standard of clinical-evidence synthesis; "low to moderate evidence quality" is not the same as "evidence-based" the way the phrase is used in marketing.

Long-term studies measuring quality-of-life outcomes are rare and often not flattering. Some newer work is starting to measure these endpoints; results so far suggest that intensive early ABA is not consistently associated with better adult quality of life, autonomy, or wellbeing.

The 2019 Sandoval-Norton paper in *Cogent Psychology* argued that ABA can be considered abuse under standard psychological definitions. The paper is controversial in the field; it is also widely cited by autistic adults as articulating something they had felt but not formalised.

If you're considering ABA Some families do choose ABA, often because it is what insurance funds, what schools recommend, or what is locally available. If you're considering it, the questions that matter most:

• What is the explicit goal of the programme? If the goal is "indistinguishable from peers" or "reduce stimming," that is a flag.
• How does the practitioner respond when a child resists, shuts down, or cries? A good answer involves stopping and reducing demand. A bad answer involves "pushing through."
• Is stimming targeted for reduction, or accepted? Soira's editorial position: stimming should be accepted.
• Does the child have meaningful choice and voice in the sessions? Even very young children can express preference. A programme that treats their preferences as relevant is different from one that doesn't.
• Are autistic adults part of the provider's training or feedback structure? Increasing numbers of providers consult autistic adults; the providers that don't are not engaging with the most important critique.
• What's the alternative if my child doesn't seem to be tolerating it? A provider that has flexibility is different from one that doesn't.

Alternatives For families uncomfortable with ABA, there are alternatives — many of which have research support of their own:

• NDBIs (naturalistic developmental behavioural interventions) — ESDM, JASPER, PRT. Blend developmental and behavioural principles in play-based, child-led ways. Increasingly the standard for early intervention research.
• DIR/Floortime — relationship-based, follows the child's lead, no behavioural-objective curriculum. Smaller evidence base but a growing one.
• Speech-language therapy focused on communication broadly (including AAC, gestalt processing, presuming competence) rather than on speech production.
• Occupational therapy with a neurodiversity-affirming sensory-integration approach.
• Parent coaching rather than direct child intervention. Empowers parents as the consistent regulators in the child's life.

None of these is universally right. The match between provider and child matters at least as much as the brand of approach.

Soira's stance We do not endorse or reject ABA. We do think the autistic-adult perspective deserves a seat at the table when a family considers it. We think provider quality matters more than the label on the door. And we think any approach whose first goal is "make the child appear less autistic" is starting in the wrong place.