5,279 Americans Were Asked What to Call Autism. There Was No Consensus.

Every intake packet, progress note and parent letter makes a choice about language. The evidence on what families actually prefer has been contradictory, and a study published in Autism on August 25 argues that is because previous work looked at averages.

The sample

A US sample of 3,181 non-autistic parents and 2,098 autistic people, of whom 451 were autistic parents. Participants rated terms across a deliberately wide span: terms for autism itself, for possible subgroups within autism, for autistic people, for autism likelihood, and for autistic characteristics.

The finding, and why it is not the obvious one

Comparing groups produced small differences. Non-autistic parents, autistic parents and autistic non-parents rated terms differently, but the authors describe the differences as “generally modest.” Taken alone, that reads as rough agreement.

Clustering the response patterns told a different story. Three groups emerged:

• The Open Cluster — almost half of all participants — “characterised by liking or neutrality towards most terms.”
• The Identity-First Cluster — support for identity-first language, strong opposition to terms for possible autism subgroups, and rejection of many pathologising terms.
• The Person-First Cluster — support for person-first language, and still some scepticism towards subgroup terms.

Autistic participants were overrepresented in the Identity-First Cluster; non-autistic parents in the Person-First Cluster. The authors' conclusion is stated plainly: “there is presently no consensus regarding autism terminology preferences in the United States.”

The turn is in the first cluster. The largest single group is not on either side of the identity-first debate — it is the one that does not mind much. A practice that picks a house style and applies it consistently is making a choice that roughly half its families will not notice, and that the other half will read as a position.

The limits

This is a survey of stated preference, not of how people respond when a term is used about their own child in a clinical document. Clustering is an exploratory technique: the three groups are a description of these responses, not categories that exist independently of the analysis. And the sample is a US convenience sample recruited for a study about terminology, which is a population already thinking about the question.

The study does not test whether terminology affects engagement, satisfaction or outcomes. It establishes that disagreement is real and structured, not that any particular choice does harm.

What you must know or do

• Ask, once, at intake, and write the answer where the clinical team will see it. One line — how the family prefers their child described — costs nothing and settles the question for that family rather than for the field. That is the only finding here that converts into practice.
• Do not treat the identity-first / person-first split as a majority question. The largest cluster is neither. Handing your team a mandated house style answers a question most families are not asking, and gets it wrong for whichever camp you did not pick.
• Look at your subgroup language specifically. Terms for possible subgroups within autism drew opposition from both the identity-first and person-first clusters — the one place the two camps in this study agreed. If your reports use severity or functioning labels as shorthand, that is the language with the least support behind it.