Error-correction problems in autism: evidence for a monitoring impairment?
Autistic kids often miss their own mistakes, so build in visible feedback and teacher-led correction every trial.
01Research in Context
What this study did
The team asked kids to launch a toy missile at a target.
Some kids had autism, some did not.
The game let children see when shots missed, yet many autistic kids kept firing without fixing their aim.
What they found
Children with autism rarely corrected visible misses.
They also failed to fix hidden errors when the toy gave no feedback.
The pattern points to a broad self-monitoring gap, not just a feedback problem.
How this fits with other research
Simpson et al. (2025) saw the same gap on a card-sort game.
Eye-tracking showed autistic teens looked away quickly after mistakes, backing the idea of weak error monitoring.
McGarty et al. (2018) and Maras et al. (2020) seem to disagree.
They found autistic adults could judge their own memory accuracy just fine.
The clash fades when you notice the tasks: adults rated feelings about pictures or words, while kids had to act and fix real-time errors.
Monitoring talk-about-memory is easier than monitoring do-in-the-moment.
Micai et al. (2019) and Feldman et al. (1999) extend the picture.
Autistic readers skip second looks at text errors and kids misjudge when they are ready to recall.
Together the studies show a domain-wide monitoring bend that starts in childhood and narrows with age.
Why it matters
For your next teaching session, do not assume a child sees and will self-repair mistakes.
Watch each trial and give an immediate, clear correction if the response is off.
Build in extra practice where the learner must check and adjust before moving on.
Over time, fade your cues so the skill of spotting and fixing errors grows from outside help to inside habit.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After any error, block the materials, re-present the cue, and physically guide the correct response before the next trial.
02At a glance
03Original abstract
With a task involving the launching of missiles at targets, Malenka et al. (1982) and C. Frith and Done (1989) showed that schizophrenic patients with delusions of alien control and auditory hallucinations were likely to leave erroneous responses uncorrected whose outcomes were not visible until the missile hit or failed to hit the target, while being able to correct visible errors adequately. This is consistent with an impairment in the central monitoring of action. Using a similar task, we found that children with autism were more likely than controls to fail to correct both kinds of error. Data are interpreted in terms of difficulties with constructing visual schemata for actions.
Journal of autism and developmental disorders, 1998 · doi:10.1023/a:1026009203333