Effect of an Artificial Intelligence-Based Online Tutor on Student Reading Growth

The current study examined the effects of an artificial intelligence (AI)–based online tutoring program on first- and second-grade students’ reading growth. Participants were 95 first- or second-grade students from one school. Reading progress was measured using fall and winter oral reading fluency (ORF) scores. Students were grouped based on program dosage (High, Low, and No Dosage Groups) using logged minutes and sessions. Multilevel analyses were used to evaluate the effects of dosage while controlling for fall reading performance and free or reduced-price lunch status. Result indicated that students in the High Dosage Group demonstrated significantly greater growth than Low Dosage or No Dosage Groups with moderate to large effects. Among students below the 25th percentile at baseline, the High Dosage Group showed significantly stronger gains than the No Dosage Group. Findings suggest AI-based tutoring may positively support early reading development, particularly for struggling readers.