To cite this article: Fajardo, I., & Joseph, H. (2024). Brief report: Autistic students read between lines. Journal of Autism and Developmental Disorders, 54, 2055–2059. https://doi.org/10.1007/s10803-022-05648-2

This is how people with ASD “read between the lines”: what the eyes reveal

Study context

Reading comprehension often hinges on inferences that connect ideas across sentences. In autism, comprehension difficulties have been linked either to challenges building those bridges or to language delays rather than autism per se. We set out to observe when and how these inferences arise during reading by preliminarily validating an English eye-tracking task.

What we investigated

We examined whether autistic students generate coherent inferences and detect inconsistencies online while reading, and where in the text such signals emerge in their eye movements.

How we did it

We tested 12 autistic students (10–15 years). They read 30 short passages in three versions: consistent, inconsistent, and control (no inference required). Classic example: seeing dark clouds then opening an umbrella (consistent) versus seeing a clear blue sky then opening an umbrella (inconsistent). We collected comprehension responses and simple eye-movement metrics: first-pass, total time, re-reading, and regressions.

What we found

• High accuracy: about 88% in inference conditions vs 99% in control—students did infer, though it was more demanding.

• Subtle eye-movement effects: only an unexpected longer gaze in a post-target region for the control condition reached significance. Descriptively, inconsistent passages drew somewhat longer times and more regressions, but these were not significant, likely due to the small sample.

Why it matters

• Psychology & Education: this challenges a blanket “deficit” view. Autistic students can make bridging inferences in brief texts; the focus should be on how they do so (e.g., strategic re-reading) and on allowing more time to connect ideas. Classroom practice can model inferences, make cause–effect links explicit, and encourage coherence monitoring (“Does this make sense?”).

• Speech-Language Therapy: combining language enrichment (vocabulary, world knowledge) with inference-building and monitoring strategies can support meaning construction when language skills are still developing.

Future directions

This was a pilot with 12 participants, limiting power and preventing inclusion of age and language skills as covariates. Larger, multilingual replications and comparisons with typically developing readers are needed. The authors share materials, data, and code to foster collaborative datasets that clarify who benefits from which supports for inference making during reading.