Understanding Variability in Reading Comprehension in Adolescents With Autism Spectrum Disorders: Interactions With Language Status and Decoding Skill

Abstract

Although it is well recognized that reading skills vary in people with autism spectrum disorders (ASD), reasons for this variability are not well understood. We used the simple view of reading model to investigate both word decoding and text comprehension processes in two well-established subtypes within the autism spectrum, those with age-appropriate structural language skills and those structural language impairments. Generally, participants with language impairments performed less well than those with age-appropriate language skills. Word-level reading was a relative strength for both groups, although it showed declines with age. Comprehension weaknesses were especially marked among those with poor structural language skills. Reading outcomes in ASD are related to variations both in decoding and comprehension and in the oral language skills that support the development of these processes.

ACKNOWLEDGMENTS

This work was funded by a Nuffield New Career Development Fellowship to C. Norbury. We are grateful to Jon Brock and Shiri Einav for their assistance with data collection. We extend a warm thank-you to Chipping Norton, Moor House, St. Dominic's, St. Catherine's, and St. Birinus schools for hosting this research. We are indebted to the young people and their families for their participation in this study.

Notes

¹These participants overlap with Norbury et al. (2009). However, one participant with ALI did not participate in this study due to school absence on the day reading measures were administered.

²We also tested the four models with nonverbal ability included at the first step, before word-level reading was entered. This did not significantly improve the fit of the model, and nonverbal ability did not predict unique variance in any of the models. Therefore, for the sake of brevity and in keeping with the aims of the article, we have reported only those models with word-level reading and oral language skill as predictor variables.