Analysis of Autistic Adolescents' Essays Using Computer Techniques.

Area of Science:

• Linguistics
• Developmental Psychology
• Computational Linguistics

Background:

• Narrative discourse challenges are common in autism spectrum disorder (ASD).
• Understanding written narratives in autistic individuals is crucial for educational and diagnostic purposes.

Purpose of the Study:

• To investigate narrative writing in autistic and non-autistic adolescents using quantitative computational methods.
• To compare written narratives of autistic and non-autistic students.

Main Methods:

• Natural Language Processing (NLP) techniques were used.
• Analysis of 333 essays from eighth-grade students (195 autistic, 138 non-autistic).

Main Results:

• Autistic students used less positive emotional language and wrote less abstract stories.
• Autistic adolescents' narratives demonstrated higher readability and complexity.
• No significant difference in teacher-assessed writing competencies between groups.

Conclusions:

• Written narratives of autistic individuals share characteristics with spoken narratives identified by computational analysis.
• National exam data could support large-scale, cost-effective autism epidemiological studies.