An Investigation Of The Impact Of Using Artificial Intelligence To Adapt Grade-level Text On The Reading Comprehension Of Middle And High School Students With Intellectual Disabilities

open_access
Open Access
Author:
Redmon, Sharon
Graduate Program:
Special Education
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
July 06, 2026
Committee Members:
Jessica Caron, Outside Unit & Field Member
Jonte Taylor, Major Field Member
David Mcnaughton, Chair & Dissertation Advisor
David Lee, Major Field Member
Mildred Boveda, Professor in Charge/Director of Graduate Studies
Keywords:
artificial intelligence
text adaptation
content area knowledge
science
social studies
intellectual and developmental disabilities
K-12 public school
middle school
high school
transition-aged students
Abstract:
This study examined the effects of the Artificial Intelligence in Reading (AIR) Prompt Strategy adapted text intervention package on the reading comprehension of content-area text with grade-level science and social studies texts by middle and high school students with intellectual and developmental disabilities (IDD). Student performance was compared using a within-subjects, repeated-measures group design, which assessed structured written recall on both Unadapted grade-level passages and passages adapted with the AIR Prompt Suite. Student participants were assigned to one of two studies based on their independent Lexile level: Study 1 (Lexile 100 and below, n = 5) and Study 2 (Lexile Level >210 and above, n = 11). No significant effect for acquisition of content-area knowledge was observed for Study 1 participants. The results demonstrated, however, that the AI-adapted text intervention package significantly increased reading comprehension of content-area text for Study 2 participants, with a statistically significant moderate effect of condition (p = .016). Efficiency data indicated that once the AIR Prompt Suite was created, it produced individualized, adapted passages and matched assessments in under four minutes per content-area text passage. In addition, a majority of both student and teacher participants reported that the AI-adapted texts were preferred to the Unadapted text. Students noted both reduced stress and increased independence, while teachers commented positively on the interventions’ goals, impact, and feasibility. The findings are discussed in terms of intervention effect, efficiency, and social validity. Limitations and implications for future research are also addressed.