Children with Autism Spectrum Disorder's Lying is Correlated with Their Working Memory But Not Theory of Mind.

The team gave children a temptation task. If kids peeked at a toy they were told not to touch, they had a chance to lie.

The sample included children with autism, children with intellectual disability, and typically developing peers.

Each child also took a working-memory game and a theory-of-mind test. The goal was to see which skill lined up with lying.

Children with autism lied less often than both other groups.

Lying frequency went hand in hand with working-memory scores, not with theory-of-mind scores.

Even kids with strong theory-of-mind skills did not lie more if their working memory was weak.

Keintz et al. (2011) first showed that autistic kids can lie but struggle to keep the story straight. Weina et al. now add that the number of lies links to working memory, not false-belief understanding.

Mazza et al. (2017) found that theory-of-mind deficits explain broader social-processing problems in autism. That looks like a clash, but the outcomes differ: Monica studied everyday social skills; Weina studied the single act of lying. Theory of mind helps with social rules, yet it does not help a child invent a quick cover story.

Eussen et al. (2016) also pinned discourse problems in autism on working-memory overload, not theory of mind. The new paper extends the same idea to deception: if memory space is tight, the lie never gets off the ground.

Stop assuming that a child who understands false beliefs will automatically lie. Check working-memory span instead. When you need a truthful report from an autistic client, reduce memory load: ask one question at a time, offer cues, or allow extra think time. If you are running social-skills groups, pair perspective-taking drills with memory games; both skills are needed but they do not grow together.

Yes. Lying is not absent in autism, but this study found it is less frequent. Using two lab tasks, a temptation-resistance task that pulls for self-protective lies (denying you peeked at a toy) and a hide-and-seek task that pulls for self-benefiting lies, autistic children told fewer lies than both typically developing children and children with intellectual disability.

In typical development, first lies are treated as a cognitive milestone. They usually emerge in the preschool years alongside executive function and theory of mind, and lying well requires holding the true state of the world in mind while asserting a false one. So a child's first lies, autistic or not, signal developing cognitive skills, not a character flaw.

The study's most useful finding is about mechanism. In typically developing children, deception is generally linked to theory of mind, the ability to represent what another person believes. In the autistic children here, lying correlated with working memory instead, and not with theory of mind. Autistic children who lied appeared to be doing it through a different cognitive route: managing the memory load of the false statement rather than modeling the listener's beliefs.

Reframe blunt honesty. Much of what reads as rudeness or social error in autistic children, like announcing that a gift is disappointing, is the flip side of telling fewer strategic lies, including the white lies neurotypical children learn to deploy early. That is a difference in social-cognitive processing, not defiance.

If an autistic child starts lying, read it developmentally before reading it morally. Given the working-memory link, a new lie can reflect growing cognitive capacity. It is also worth asking what the lie gets the child: escape from a demand or avoidance of a consequence are functions a BCBA can assess and address the same way as any other behavior, by changing what honesty and disclosure produce.

For clinicians, two practical notes. First, do not assume an autistic client cannot lie during assessment or self-report; this research shows deception exists in the repertoire. Second, if perspective-taking programs are on the treatment plan, this study cautions against assuming theory-of-mind training will change deception-related behavior, because in this population the two were not correlated.